WO2009104460A1

WO2009104460A1 - Fatigue analysis device and computer program

Info

Publication number: WO2009104460A1
Application number: PCT/JP2009/051366
Authority: WO
Inventors: 悦則藤田; 直輝落合; 重行小島; 由美小倉
Original assignee: 株式会社デルタツーリング
Priority date: 2008-02-20
Filing date: 2009-01-28
Publication date: 2009-08-27
Also published as: JP2009195384A; JP5086834B2

Abstract

A fatigue degree is obtained by a sense evaluation value considering compensation by a sympathetic nerve activity. Provided is a fatigue analysis device including: first fatigue degree calculation means (11) which obtains a reference fatigue degree; second fatigue degree calculation means (12) which obtains a corrected fatigue degree by using the maximum Lyapunov exponent; and judgment means (13) which judges whether a fatigue compensation by the sympathetic nerve activity is present. In the time band when it is judged that no fatigue compensation is performed by the sympathetic nerve activity, accumulated fatigue degree output means (14) outputs an accumulated sum of the reference fatigue degrees obtained by the first fatigue degree calculation means (11). In the time band when it is judged that a fatigue compensation is performed by the sympathetic nerve activity, accumulated fatigue degree output means (14) outputs an accumulated sum of the corrected fatigue degrees obtained by the second fatigue degree calculation means (12). Since the reference fatigue degree or the corrected fatigue degree is outputted in accordance with the presence/absence of the fatigue compensation by the sympathetic nerve activity, the accumulated fatigue degree approaches the sense evaluation value.

Description

Fatigue analysis apparatus and computer program

The present invention relates to a fatigue analysis apparatus for obtaining a degree of fatigue as a degree of human fatigue and a computer program used therefor.

Patent Document 1 describes the degree of fatigue of a person by comparing the amount of energy by the work of muscles used for maintaining the homeostasis taken by a person with fatigue by the amount of metabolism of the product by the work. The technique calculated | required as the fatigue degree used as is disclosed. This technology collects human biological signals, calculates the difference between the peak value on the upper limit side and the peak value on the lower limit side for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data, Power value calculating means for calculating the difference as a power value, power value inclination calculating means for calculating the inclination of the power value with respect to the time axis in a predetermined time range by slide calculation a predetermined number of times, and the inclination of the obtained power value And a means for calculating an integral value by processing an absolute value of the time series signal, and the obtained integral value is obtained as a fatigue level.
WO2005 / 039415A1 publication

However, for example, when comparing the degree of fatigue obtained by the technique disclosed in Patent Document 1 and the sensory evaluation value of muscle fatigue by the Borg scale measured at the same time when the person continues to sit on the seat, It turned out that both may not show a similar tendency. When the present inventors examined this point, anti-gravity muscles (posture muscles) used for maintaining the sitting posture with respect to the state in which the fatigue level required by Patent Document 1 maintains and continues the sitting posture. While only looking at metabolism associated with muscle contraction, such as energy consumption of the body and energy consumption due to sudden muscle contraction due to the occurrence of pain associated with continued sitting posture, I thought that not only fatigue but also mental fatigue was considered. In other words, the technique of Patent Document 1 is based on the premise of seeking fatigue that occurs in a mentally relaxed state, and if the fatigue is covered (compensated) by the sympathetic nerve function and does not appear as muscle contraction, the sympathetic nerve activity Even if the blood flow changes due to, it is not considered as fatigue level. However, in reality, sympathetic nerve activity caused mental fatigue and accompanying changes in blood flow, and this was considered to be the main cause of the case where a deviation from the sensory evaluation value occurred.

The present invention has been made in view of the above, and it is an object of the present invention to provide a fatigue analysis apparatus and a computer program capable of obtaining a fatigue level closer to a sensory evaluation value in consideration of a metabolic amount of compensation due to sympathetic activity. To do.

In order to solve the above-described problems, a fatigue analysis apparatus according to the present invention is a fatigue analysis apparatus that performs fatigue analysis using pulse wave biosignal data collected by a biosignal measuring device, and is an original waveform of the biosignal data The difference between the peak value on the upper limit side and the peak value on the lower limit side is calculated for each predetermined time range from the peak value of each cycle, and this difference is used as the power value. Based on the time series data of the power value, the standard fatigue is calculated. Obtained by correcting the time series data of the power value used in the first fatigue degree calculating means using the first fatigue degree calculating means for obtaining the degree and the maximum Lyapunov exponent obtained from the biological signal data. A second fatigue degree calculating means for obtaining a corrected fatigue level based on the corrected power value time-series data, a determination means for determining whether or not there is a compensation effect of fatigue due to the sympathetic nerve activity, and a sympathetic nerve activity by the determination means Fatigue due to In a time zone in which it is determined that no compensation has been made, a cumulative sum of the reference fatigue levels obtained by the first fatigue level calculation means is obtained and output, and a state in which compensation for fatigue due to sympathetic nerve activity is being made The determined time zone includes a cumulative fatigue level output unit that calculates and outputs a cumulative sum of the corrected fatigue levels obtained by the second fatigue level calculation unit.

The determination means obtains a maximum Lyapunov exponent from the vital sign data, and calculates a degree of arousal based on time series data of the maximum Lyapunov exponent, and arousal degree obtained by the arousal degree calculation means , And a timing determination unit that determines a timing at which the corrected fatigue level of the second fatigue level calculation unit is adopted, and the cumulative fatigue level output unit includes the second fatigue level calculation unit by the timing determination unit. In a time zone before it is determined that the corrected fatigue level is adopted, a cumulative sum of the reference fatigue levels obtained by the first fatigue level calculation unit is obtained and output, and the corrected fatigue level of the second fatigue level calculation unit is calculated. In the time zone after it is determined that the degree is adopted, the cumulative fatigue correction degree obtained by the second fatigue degree calculating means is calculated and output. Door is preferable.

Further, the first fatigue level calculating means calculates a difference between the peak value on the upper limit side and the peak value on the lower limit side for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data, A power value calculating means for calculating a time series data of the power value using the difference as the power value; a power value inclination calculating means for obtaining a slope of the power value with respect to a time axis in a predetermined time range by performing slide calculation a predetermined number of times; A power value slope integrating means for performing absolute value processing on a time series signal of the power value slope obtained by the slide calculation by the power value slope calculating means, and calculating an integral value for each predetermined time range as the reference fatigue level; It is preferable to comprise.

Furthermore, it comprises LF / HF time series data calculating means for obtaining time series data of the power spectrum of LF / HF obtained by frequency analysis of time series data of heart rate or pulse rate collected by the biological signal measuring instrument, The power value calculation means multiplies the value of the time series data of the power spectrum of the LF / HF and the value of the time series data of the power value by the values corresponding to each other and LF / HF It is preferably set to calculate the time series data of the LF / HF included power value as the calculated power value.

Further, the second fatigue level calculating means includes the value of the time series data of the maximum Lyapunov exponent used in the arousal level calculating means and the value of the time series data of the power value used in the first fatigue level calculating means. A correction power value calculating means for calculating the time series data of the correction power value by multiplying the values corresponding to each other at the corresponding time, and the inclination of the correction power value with respect to the time axis in a predetermined time range Correction power value inclination calculation means obtained by sliding calculation a predetermined number of times, and time-series signals of the correction power value inclination obtained by slide calculation by the correction power value calculation means are subjected to absolute value processing for each predetermined time range. The correction power value gradient integration means for calculating the integrated value as the corrected fatigue degree can be provided.

Further, the second fatigue level calculating means includes the time series data value of the maximum Lyapunov exponent used in the arousal level calculating means and the time series data of the LF / HF included power value used in the first fatigue level calculating means. And a correction power value calculation means for calculating time series data of the correction power value by multiplying the values of the correction values by values at times corresponding to each other, and a correction power value in a predetermined time range of the correction power value A correction power value inclination calculating means for obtaining the inclination with respect to the time axis by slide calculation a predetermined number of times, and a time series signal of the inclination of the correction power value obtained by the slide calculation by the correction power value calculation means are subjected to absolute value processing, A correction power value gradient integration means for calculating an integrated value for each predetermined time range as the corrected fatigue level may be provided.

The arousal degree calculating means calculates a maximum Lyapunov exponent from the vital sign data and calculates time series data of the maximum Lyapunov exponent, and a slope of the maximum Lyapunov exponent with respect to a time axis in a predetermined time range. The maximum Lyapunov exponent slope calculating means obtained by sliding calculation a predetermined number of times, and the time series signal of the slope of the maximum Lyapunov exponent slope obtained by the slide calculation by the maximum Lyapunov exponent slope calculating means is subjected to absolute value processing, and a predetermined time range. A maximum Lyapunov exponent slope integration means for calculating the integral value for each arousal level, and a cumulative arousal level for obtaining and outputting a cumulative sum of the arousal levels obtained by the maximum Lyapunov exponent slope integration means Output means.

In addition, the timing determination unit is configured to determine the sudden conversion point of the slope of the cumulative arousal level output from the cumulative arousal level output unit as a timing to employ the corrected fatigue level of the second fatigue level calculation unit. It is preferable.

Further, the timing determination means obtains the slope of the cumulative arousal degree in a predetermined time range from the start of measurement as a reference slope, and the slope of the cumulative arousal degree in the predetermined time range in the time zone after obtaining the reference slope is a reference It is preferable that the abrupt conversion point is determined when a change of at least 10% occurs with respect to the inclination.

The computer program of the present invention is a computer program introduced into a fatigue analysis apparatus that performs fatigue analysis using pulse wave biosignal data collected by a biosignal measuring device, and the original waveform of the biosignal data The difference between the peak value on the upper limit side and the peak value on the lower limit side is calculated for each predetermined time range from the peak value of each cycle, and this difference is used as the power value. Based on the time series data of the power value, the standard fatigue is calculated. Obtained by correcting the time series data of the power value used in the first fatigue degree calculating means using the first fatigue degree calculating means for obtaining the degree and the maximum Lyapunov exponent obtained from the biological signal data. A second fatigue level calculating means for obtaining a corrected fatigue level based on the corrected power value time-series data, a determination means for determining the presence or absence of a compensation effect of fatigue due to sympathetic nerve activity, and the determination means In a time zone in which it is determined that the fatigue due to the sympathetic nerve activity is not compensated, the cumulative sum of the reference fatigue levels obtained by the first fatigue level calculating means is obtained and output, and the fatigue level due to the sympathetic nerve activity is calculated. And a cumulative fatigue level output means for obtaining and outputting a cumulative sum of the corrected fatigue levels obtained by the second fatigue level calculation means in a time zone determined to be in a state of being compensated. .

Further, the second fatigue level calculating means calculates the value of the time series data of the maximum Lyapunov exponent used by the arousal level calculating means and the value of the time series data of the power value used by the first fatigue level calculating means. A correction power value calculating means for calculating the time series data of the correction power value by multiplying the values corresponding to each other at the corresponding time, and the inclination of the correction power value with respect to the time axis in a predetermined time range Correction power value inclination calculation means obtained by sliding calculation a predetermined number of times, and time-series signals of the correction power value inclination obtained by slide calculation by the correction power value calculation means are subjected to absolute value processing for each predetermined time range. The correction power value gradient integration means for calculating the integrated value as the corrected fatigue degree can be provided.

According to the present invention, based on time-series data of power values, or based on values (LF / HF included power values) in consideration of time-series data of LF / HF power spectrum in time-series data of power values. A time series of the power value used in the first fatigue degree calculating means or the LF / HF power value using the first fatigue degree calculating means for obtaining the reference fatigue degree and the maximum Lyapunov exponent obtained from the biological signal data A second fatigue level calculating means for correcting the data value and obtaining a corrected fatigue level based on the obtained time series data of the corrected power value; a determining means for determining the presence or absence of a compensation effect of fatigue due to sympathetic nerve activity; The cumulative fatigue level output means calculates and outputs the cumulative sum of the standard fatigue levels obtained by the first fatigue level calculation means in a time zone when it is determined that the fatigue due to sympathetic nerve activity is not compensated. And In the time zone in which the price of fatigue is determined to state that made by neural activity, a structure that obtains and outputs the cumulative sum of the corrected fatigue obtained by the second fatigue degree calculating unit. That is, depending on whether or not there is compensation for fatigue due to sympathetic nerve activity, the fatigue level based on only the power value or only the LF / HF power value (reference fatigue level) or the fatigue level considering the maximum Lyapunov exponent (correction) Therefore, the change in the cumulative fatigue level is closer to the sensory evaluation value, that is, closer to the actual fatigue feeling.

The determination means obtains the maximum Lyapunov exponent from the biological signal data, calculates the degree of arousal based on the time series data of the maximum Lyapunov exponent, and the second degree from the arousal degree obtained by the arousal degree calculator. It is preferable that the apparatus includes a timing determination unit that determines a timing at which the corrected fatigue level of the fatigue level calculation unit is adopted. The higher the maximum Lyapunov exponent, the higher the adaptability, tension and concentration to things, and the lower the maximum Lyapunov exponent, the more mentally relaxed. The “arousal degree” obtained by the determination means employed in the present invention is set so that the slope of the maximum Lyapunov exponent is obtained from the time series data of the maximum Lyapunov exponent, and the slope is obtained by absolute value processing and integration. is doing. As a result, the “degree of arousal” captures a global trend of how the maximum Lyapunov exponent, which indicates the degree of stimulation received mentally, fluctuates. In other words, “arousal level” is an index that captures fatigue that cannot be captured as muscle contraction, that is, mental fatigue of the brain that cannot be captured as muscle fatigue, because it is compensated for by sympathetic nerve tension. (Ergonomics Vol. 40, No. 5 (2004), “Development of a simple evaluation method for long-term sitting fatigue using fingertip plethysmogram information” (Yasunori Fujita, etc.)). Therefore, while the compensation for fatigue due to sympathetic activity is performed, the fatigue level (reference fatigue level) is calculated without considering the arousal level, while the compensation for fatigue due to sympathetic activity is calculated from the value of the arousal level. If it is determined that it has been carried out, a corrected power value incorporating the maximum Lyapunov exponent is obtained, and the fatigue level (corrected fatigue level) is obtained again using the corrected power value. It becomes closer to the evaluation value.

Strictly speaking, mental fatigue includes the mental fatigue of the autonomic nervous system and the mental fatigue of the brain, but the mental fatigue of the autonomic nervous system is similar to the mental fatigue of the brain. It becomes the object of compensation by the activity. For example, when a fatigue analysis is performed after lying down and taking a sufficient rest or sleep, there is almost no physical fatigue of antigravity muscles at the start of the analysis, and the autonomic nervous system and brain mental fatigue None of this has occurred so much. For this reason, when determining the fatigue level from this state, first determine the standard fatigue level by peripheral physical fatigue, and then determine the fatigue level considering the mental fatigue of the brain based on the arousal level, It approaches human sensory evaluation values. However, for example, when measuring the degree of fatigue while sitting in a seat during daily activities, it occurs with physical fatigue of the peripheral system at a stage before the mental fatigue of the brain is greatly affected. There is also mental fatigue of the autonomic nervous system. Therefore, in calculating the fatigue level before starting to consider mental fatigue of the brain, that is, the standard fatigue level, the peak value of each cycle of the original waveform of the biological signal data, which is an index of peripheral physical fatigue, is used. The LF / HF obtained from the heart rate or pulse rate, which is an index of mental fatigue of the autonomic nervous system, is calculated as the difference between the peak value on the upper limit side and the peak value on the lower limit side for each predetermined time range. It is preferable to adopt a configuration in which a value considering the time series data of the power spectrum (LF / HF included power value) is employed, thereby obtaining the reference fatigue level.

According to the present invention, since the fatigue level close to the actual fatigue feeling can be obtained, the obtained fatigue level can be used for various applications. For example, it is possible to more accurately determine the seat and bedding that are suitable for a person based on the degree of fatigue when the person is seated on the seat or the degree of fatigue when the person is lying on the bedding. This is because whether the seat is good or bad is determined by the feeling of sitting fatigue, and whether the bedding is good or bad is also determined by the feeling of fatigue when lying down.

FIG. 1 is a block diagram showing a configuration of a fatigue analysis apparatus according to an embodiment of the present invention. FIG. 2 is a diagram for explaining the configuration of the first fatigue level calculation means, the second fatigue level calculation means, and the cumulative fatigue level output means introduced into the fatigue analysis apparatus. FIG. 3 is a diagram for explaining a configuration of a determination unit introduced into the fatigue analysis apparatus. FIG. 4 is data showing an example of heart rate fluctuations collected from an electrocardiograph for calculating LF / HF time-series data. FIG. 5 is a diagram showing the power spectra of the HF component and the LF component obtained by continuous wavelet analysis of the time-series data of the heart rate shown in FIG. FIG. 6 shows LF / HF time-series data obtained from FIG. FIG. 7 shows time-series data of average values calculated every 5 seconds from the LF / HF time-series data of FIG. FIG. 8 is a diagram for explaining the LF / HF inclusion means, in which FIG. 8A is a diagram obtained by multiplying the time series data of FIG. 7 by 1/5, and FIG. FIG. 8C shows time series data of average values of power values every 5 seconds obtained by processing time series data of pulse waves obtained by the volume pulse wave meter. FIG. This is time series data of LF / HF power values obtained by multiplying the series data and the time series data of FIG. 8B by corresponding time values. FIG. 9 is a diagram illustrating an example of time-series data of the average value of the maximum Lyapunov exponent used for calculation of the corrected power value every 5 seconds. FIG. 10 is a diagram for explaining the process of calculating the maximum Lyapunov exponent. FIG. 10A is an example of time-series data of the average value of the maximum Lyapunov exponent every 5 seconds (measurement time 1800 seconds). FIG. 10B shows the average value of the power values every 5 seconds obtained by processing the time-series data of the pulse wave obtained by the fingertip plethysmograph by the power value calculating means. FIG. 10C is an example of time-series data, and FIG. 10C is a correction obtained by multiplying the time-series data of FIG. 10A and the time-series data of FIG. 10B by corresponding time values. It is an example of the time series data of a power value. FIG. 11 is a diagram showing test results when test subject A sits on a urethane sheet under static conditions in Test Example 1, (a) shows a time-series change in the degree of subjective fatigue, and (b) shows FIG. 4 is a diagram showing fatigue curves (A) to (D) together. FIG. 12 is a diagram showing test results when test subject A is seated on a wooden chair under static conditions in Test Example 1, (a) shows a time-series change in the degree of subjective fatigue, and (b) shows FIG. 4 is a diagram showing fatigue curves (A) to (D) together. FIG. 13 is a diagram showing test results when test subject A is seated on a urethane sheet in a vibration condition in Test Example 1, (a) shows a time-series change in the degree of subjective fatigue, and (b) FIG. 3 is a diagram showing fatigue curves (A) to (D) together. FIG. 14 is a diagram showing test results when subject A sits on a wooden chair under vibration conditions in Test Example 1, (a) shows time-series changes in the degree of subjective fatigue, and (b) FIG. 3 is a diagram showing fatigue curves (A) to (D) together. FIG. 15 is a diagram showing test results when test subject B sits on a wooden chair under static conditions in Test Example 1, (a) shows the time-series change in the degree of subjective fatigue, and (b) shows FIG. 4 is a diagram showing fatigue curves (A) to (D) together. FIG. 16 is a diagram showing test results when test subject A is seated on a urethane sheet in a static condition in Test Example 2, (a) shows a time-series change in the degree of subjective fatigue, and (b) shows A time series change of the arousal level / time and the cumulative arousal level is shown, and (c) is a diagram showing the fatigue curves (A) to (D) together. FIG. 17 is a diagram showing test results when subject A sits on a wooden chair under static conditions in Test Example 2, (a) shows time-series changes in the degree of subjective fatigue, and (b) A time series change of the arousal level / time and the cumulative arousal level is shown, and (c) is a diagram showing the fatigue curves (A) to (D) together. FIG. 18 is a diagram showing test results when test subject A is seated on a urethane sheet in a vibration condition in Test Example 2, (a) shows a time series change in the degree of subjective fatigue, and (b) shows arousal. A time series change of degree / time and cumulative arousal degree is shown, and (c) is a diagram showing fatigue curves (A) to (D) together. FIG. 19 is a diagram showing test results when test subject A is seated on a wooden chair under vibration conditions in Test Example 2, (a) shows time-series changes in subjective fatigue level, and (b) shows arousal. A time series change of degree / time and cumulative arousal degree is shown, and (c) is a diagram showing fatigue curves (A) to (D) together. FIG. 20 is a diagram showing test results when subject B sits on a wooden chair under static conditions in Test Example 2, (a) shows the time-series change in the degree of subjective fatigue, and (b) shows A time series change of the arousal level / time and the cumulative arousal level is shown, and (c) is a diagram showing the fatigue curves (A) to (D) together. FIG. 21 is a diagram showing test results when test subject A is seated on a urethane sheet in a static condition in Test Example 3, (a) shows a time-series change in the degree of subjective fatigue, and (b) shows It is the figure which showed each fatigue curve (A)-(D) together, (c) shows the time series change of the arousal degree / time and the cumulative arousal degree, and (d) is the time of the arousal degree / time. It is the figure which set the threshold value to the series change. FIG. 22 is a diagram showing test results when subject A sits on a wooden chair under static conditions in Test Example 3, (a) shows the time-series change in the degree of subjective fatigue, and (b) shows It is the figure which showed each fatigue curve (A)-(D) together, (c) shows the time series change of the arousal degree / time and the cumulative arousal degree, and (d) is the time of the arousal degree / time. It is the figure which set the threshold value to the series change. FIG. 23 is a diagram showing test results when test subject A is seated on a urethane sheet in a vibration condition in Test Example 3; (a) shows a time-series change in the degree of subjective fatigue; and (b) It is the figure which showed fatigue curves (A)-(D) together, (c) shows the time series change of the arousal degree / time and the cumulative arousal degree, and (d) is the time series of the arousal degree / time. It is the figure which set the threshold value to the change. FIG. 24 is a diagram showing test results when test subject A is seated on a wooden chair under vibration conditions in Test Example 3, (a) shows the time-series change in the degree of subjective fatigue, and (b) It is the figure which showed fatigue curves (A)-(D) together, (c) shows the time series change of the arousal degree / time and the cumulative arousal degree, and (d) is the time series of the arousal degree / time. It is the figure which set the threshold value to the change. FIG. 25 is a diagram showing test results when subject B sits on a wooden chair under static conditions in Test Example 3, (a) shows a time-series change in the degree of subjective fatigue, and (b) It is the figure which showed each fatigue curve (A)-(D) together, (c) shows the time series change of the arousal degree / time and the cumulative arousal degree, and (d) is the time of the arousal degree / time. It is the figure which set the threshold value to the series change.

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, the present invention will be described in more detail based on the embodiments shown in the drawings. FIG. 1 is a diagram illustrating a configuration of a fatigue analysis apparatus 1 according to the present embodiment. As shown in this figure, the fatigue analysis apparatus 1 is composed of a computer or the like, and includes data receiving means 10 for receiving data of a fingertip plethysmograph 20 as a biological signal measuring instrument, and also receives data. First fatigue level calculation means 11, second fatigue level calculation means 12, determination means 13, and cumulative fatigue level output, which are computer programs for processing time series data of fingertip volume pulse waves as biological signal data received by means 10 Means 14 is set.

The first fatigue level calculating means 11 includes a power value calculating means 111, a power value slope calculating means 112, and a power value slope integrating means 113, as shown in FIG. The power value calculation unit 111 includes the following processing steps. First, for the time series data of the fingertip plethysmogram collected by the fingertip plethysmograph 20 and received by the data receiving means 11, the maximum value and the minimum value are obtained by the smoothing differential method by Savitzky and Golay, respectively. . Next, the local maximum value and the local minimum value are divided every predetermined time range set in advance, in this embodiment, every 5 seconds, and the average value (maximum peak value) of the local maximum value and the local minimum value in the time range. The average value (peak value on the lower limit side) of the values is obtained, and the difference between them is obtained as the power value. However, in order to emphasize the amount of change, in this embodiment, the difference between the average value of the maximum value and the average value of the minimum value in the predetermined time range is squared to obtain the power value.

The power value inclination calculation means 112 obtains the inclination with respect to the time axis by the least square method for a predetermined time width Tw (180 seconds in the present embodiment) from the time series data of the power values obtained by the power value calculation means 111. Next, the next time width Tw (180 seconds) is similarly calculated with the overlap time Tl (162 seconds), and the result is plotted. This calculation (slide calculation) is repeated sequentially. Thereby, in this example, the inclination of the power value is plotted every 18 seconds, and the time series data is obtained.

For example, when obtaining the slope during T seconds (s) with an overlap rate of 90%, first, the slope of the power value between 0 (s) and T (s) with respect to the time axis is obtained by least square approximation. . Then
Slide calculation (1): Between T / 10 (s) and T + T / 10 (s),
Slide calculation (2): Between 2 × T / 10 (s) and T + 2 × T / 10 (s),
Slide calculation (n): Each slope between n × T / 10 (s) and T + n × T / 10 (s) is obtained by least square approximation.

The power value slope integration means 113 first performs absolute value processing on the time series signal of the power value slope obtained by the slide calculation by the power value slope calculation means 112. That is, the power values obtained every 18 seconds in the above example are all made positive. Next, an integral value of a time (in this example, 18 seconds) from a time point at which a slope of a certain power value is plotted to a time point at which the slope of the next power value is plotted is obtained every predetermined time range. Then, the obtained integral value (obtained every 18 seconds in this example) is set as the reference fatigue level.

Here, in the above description, the power value gradient calculating unit 112 calculates the power value gradient using the power value obtained from the power value calculating unit 111 to obtain the reference fatigue level. As described above, it is preferable to obtain the above in consideration of mental fatigue of the autonomic nervous system. Therefore, LF / HF time-series data calculation for obtaining time-series data of LF / HF power spectrum obtained by frequency analysis of time-series data of heart rate or pulse rate measured by an electrocardiograph or fingertip plethysmograph Means 111a is provided, and the power value calculation means 111 adds the value of the time series data of the LF / HF power spectrum obtained by the LF / HF time series data calculation means 111a to the value of the time series data of the power value. It is preferable to obtain the LF / HF power value.

The method of adding the value of the time series data of the power spectrum of LF / HF to the power value is as follows. As a frequency analysis method, it is preferable to use continuous wavelet transform. When the frequency analysis of the heart rate or pulse rate is performed, the LF component is 0.04 to 0.15 Hz and the HF component is 0.15 to 0.4 Hz. However, since the continuous wavelet transform has a high resolution, Waves can be taken well and is suitable for frequency analysis of heart rate or pulse rate variation.

Specifically, first, as heart rate or pulse rate measured by an electrocardiograph or fingertip plethysmograph, time-series data of heart rate or pulse rate per minute obtained from RR interval data is obtained. Using this, continuous wavelet analysis is performed to calculate the total value of the power spectrum of the HF component and the LF component per unit time (RR interval) to obtain LF / HF time series data (FIGS. 4 to 5). 6). Next, an average value is obtained every 5 seconds for the LF / HF time-series data (see FIG. 7). Since LF / HF is obtained every RR interval, the average value every 5 seconds includes the sum of the LF / HF values for each RR interval included in 5 seconds. Calculated by dividing by the number of RR intervals. Then, the time-series data of the average value every 5 seconds is multiplied by the time-series data of the power value calculated every 5 seconds similarly (see FIG. 8). Thereby, time series data of the LF / HF inclusion power value is obtained.

The average value of LF / HF may be used as it is, but it may be multiplied by the power value after being multiplied by several times (for example, 1/5 times) for easy processing. Moreover, although the average value of LF / HF and said power value are calculated | required every 5 second, the time interval which calculates | requires this average value is an example to the last. In order to obtain the LF / HF inclusion power value by multiplying the values output in both time series, the average value may be obtained at the same time interval so that the calculation time of the values in both coincides. It is not limited to seconds.

As shown in FIG. 2, the second fatigue level calculating means 12 includes a corrected power value calculating means 121, a corrected power value slope calculating means 122, and a corrected power value slope integrating means 123. The corrected power value calculating means 121 obtains the time series data of the maximum Lyapunov exponent from the time series data of the fingertip volume pulse wave, the value of the time series data of the maximum Lyapunov exponent, and the power of the first fatigue level calculating means 11 The power value time-series data values used by the value calculating means 111 are multiplied by the values corresponding to each other in time to obtain a corrected power value, and the corrected power value time-series data is calculated.

Specifically, the value of the maximum Lyapunov exponent is obtained every second from the time series data of the fingertip volume pulse wave. Further, in this embodiment, the average value every 5 seconds is calculated from the time series data of this maximum Lyapunov exponent. Is calculated (FIGS. 9 and 10A). Then, the average value every 5 seconds is multiplied by the power value obtained every 5 seconds (FIG. 10B), and the time series data of the corrected power value including the maximum Lyapunov exponent (FIG. 10 ( c)). Since the maximum Lyapunov exponent is obtained by the arousal degree calculation means 131 described later, the second fatigue degree calculation means 12 does not obtain the maximum Lyapunov exponent again, but calculates the corrected power value of the maximum Lyapunov exponent obtained by the arousal degree calculation means 131. It is preferable to use it. However, when the first fatigue level calculating unit 11 calculates the reference fatigue level from the LF / HF included power value, when the second fatigue level calculating unit 12 starts operating, the corrected power value calculating unit 121 Instead of the time series data of / HF, the value of the time series data of the maximum Lyapunov exponent is multiplied by the value of the time series data of the power value. In this case, the corrected power value calculation means 121 multiplies the power value time series data multiplied by the LF / HF time series data and the time series data value of the maximum Lyapunov exponent as it is. It can also be used as a correction power value.

The corrected power value slope calculating means 122 has the same configuration as that of the power value slope calculating means 112 described above except that the corrected power value is used as data serving as a basis for calculation. Find the tilt with respect to the axis.

The correction power value slope integration means 123 is the same as that of the power value slope integration means 113 except that the slope of the correction power value is used as the basis data for the calculation. These time series signals are subjected to absolute value processing, and an integral value for each predetermined time range is calculated. This integrated value is used as the corrected fatigue level.

As shown in FIG. 3, the determination unit 13 includes an arousal degree calculation unit 131, a cumulative arousal level output unit 132, and a timing determination unit 133. The arousal degree calculation means 131 further includes a maximum Lyapunov exponent calculation means 131a, a maximum Lyapunov exponent slope calculation means 131b, and a maximum Lyapunov exponent slope integration means 131c. The maximum Lyapunov exponent calculation means 131a calculates the maximum Lyapunov exponent from the time series data of the fingertip plethysmogram received by the data receiving means 11, and calculates the time series data of the maximum Lyapunov exponent. Specifically, the time series data of the fingertip volume pulse wave is first reconstructed into a state space by a time delay method. The time-series delay time of the fingertip plethysmogram is 50 ms, and when the FNN (False Near Neighbors) method is used as the embedding dimension, the FNN becomes almost zero in the dimension 3 and completely zero in the dimension 4 Therefore, the optimum embedding dimension is set to 4 dimensions. Here, with respect to the obtained continuous data calculation value, the attractor is reconfigured with a time width of 30 seconds, and the value of the maximum Lyapunov exponent is plotted every second by sliding the time width by 1 second, Find time series data of the maximum Lyapunov exponent.

The maximum Lyapunov exponent slope calculating means 131b obtains the slope of the maximum Lyapunov exponent with respect to the time axis in a predetermined time range by performing slide calculation a predetermined number of times. Specifically, the maximum Lyapunov exponent determined as described above is smoothed by a smoothing differential method using Savitzky and Golay to obtain a maximum value and a minimum value, and then plots the time series data of the maximum value and the minimum value. The inclination with respect to the time axis in a predetermined time range is obtained. The method of obtaining the inclination is the same as that of the power value inclination calculating means 112 described above. First, the inclination with respect to the time axis is obtained by the least square method for the predetermined time width Tw (180 seconds in the present embodiment). Next, the next time width Tw (180 seconds) is similarly calculated with the overlap time Tl (162 seconds), and the result is plotted. This calculation (slide calculation) is repeated sequentially. Thus, in this example, the slope of the maximum Lyapunov exponent is plotted every 18 seconds, and the time series data is obtained.

The maximum Lyapunov exponent slope integration means 131c first performs absolute value processing on the time series signal of the slope of the maximum Lyapunov exponent slope obtained by the slide calculation by the maximum Lyapunov exponent slope calculation means 131b. That is, the maximum Lyapunov exponent slope obtained every 18 seconds in the above example is set to a positive value. Next, an integral value of a time (in this example, 18 seconds) from a time point when a slope of a certain maximum Lyapunov exponent is plotted to a time point when a slope of the next maximum Lyapunov exponent is plotted is obtained every predetermined time range. Then, the obtained integral value (obtained every 18 seconds in this example) is used as an arousal level (Ergonomics Vol. 40, No. 5 (2004), “Long-term sitting fatigue using fingertip volume pulse wave information” Development of a simple evaluation method "(Fujita, Akinori, etc.)). As described above, the degree of arousal indicates a tendency of fluctuation of the maximum Lyapunov exponent indicating the mental state, and thus becomes an index indicating the degree of fatigue compensated by the sympathetic nerve tension.

The cumulative arousal output means 132 calculates and outputs a cumulative sum of each integral value (arousal degree) obtained by the maximum Lyapunov exponent slope integral means 131b. Thereby, the fluctuation tendency of arousal level can be grasped.

The timing determination unit 133 uses the sudden conversion point of the slope of the cumulative evoke level output by the cumulative evoke level output unit 132 and uses the corrected fatigue level of the second fatigue level calculation unit 12 in the cumulative fatigue level output unit 14. It is a means to determine. The “sudden conversion point” is obtained by using the slope of the cumulative arousal level in the predetermined time range from the start of measurement as the reference slope, and the slope of the cumulative arousal level in the predetermined time range in the time zone after obtaining this reference slope is the reference slope On the other hand, it is preferable that the change occurs at least 10% or more, depending on the person, 20% or more to 30% or more. This is because there are individual differences in the ability of people to adjust to fatigue, and there is an acceptable range of about ± 10% innately, and an acceptable range of about ± 20% to ± 30% when considering the acquired ability of adjustment. It is based on being. For example, regarding the maintenance of the sitting posture, the adjustment ability varies depending on how to train the antigravity muscle, the breathing method, and the cardiopulmonary function. The reference inclination is preferably a time zone before mental fatigue is greatly affected, and when analyzing fatigue when seated, such as a car seat or office chair, 3 to 15 minutes after the start of measurement. It is preferable to adopt an average slope of 3 minutes to 12 minutes in the range. The slope of the cumulative arousal level compared with the reference slope is preferably an average slope for 2 minutes to 5 minutes.

The cumulative fatigue level output means 14 is a power value gradient of the first fatigue level calculation means 11 in a time zone before the timing determination means 133 determines that the corrected fatigue level of the second fatigue level calculation means 12 is adopted. The cumulative sum of the standard fatigue levels obtained by the integrating means 113 is obtained and output. On the other hand, in the time zone after it is determined that the second fatigue level calculating unit 12 is adopted, the cumulative sum of the corrected fatigue levels obtained by the corrected power value slope integrating unit 123 of the second fatigue level calculating unit 12 is obtained. Output. As a result, the accumulation of fatigue that occurs in the relaxed state is obtained as the cumulative sum of the standard fatigue levels, and the accumulation of fatigue in the state where the mental fatigue compensated by the exchange nerve activity is taken into account is the cumulative sum of the corrected fatigue levels. As required. As a result, the change in the cumulative fatigue level becomes closer to the sensory evaluation value as compared with the case based on the fatigue level calculated only by the conventional power value. However, the determination of the sudden conversion point of the inclination by the timing determination unit 133 compares the average inclinations of 2 minutes to 5 minutes, so that the occurrence of the sudden conversion point is 2 minutes from the actual generation time of the sudden conversion point. After 5 minutes. After the determination, the cumulative fatigue level output means 14 outputs the cumulative sum of the corrected fatigue levels of the second fatigue level calculation means 12.

As for the calculation of the corrected fatigue level, it is preferable to automatically determine the timing at which the corrected fatigue level is adopted using the timing determination unit 133 as in the fatigue analysis apparatus 1 of the present embodiment, but from 3 minutes to 15 minutes after the start of measurement. A predetermined threshold value is set from the “degree of arousal / time” (every 18 seconds in this embodiment) obtained by the maximum Lyapunov exponent slope integration means 131c in the range, and when the predetermined threshold value is exceeded, the second fatigue level is calculated. It can also be set so that the cumulative sum of the corrected fatigue levels of the means 12 is output. In addition, the excitation suppression coexistence period (ergonomics Vol. 40, No.5 (2004) “fingertips) appearing in opposite phase when time series data of power value slope and time series data of maximum Lyapunov exponent slope are overlaid and displayed. "Development of a simple evaluation method for long-term sitting fatigue using volume pulse wave information" (see Yasunori Fujita, etc.) can also be determined as the timing for outputting the corrected fatigue level.

Note that a computer program including the first fatigue level calculation unit 11, the second fatigue level calculation unit 12, the determination unit 13, the cumulative fatigue level output unit 14, and the like is provided by being stored in a recording medium. . A “recording medium” is a medium that can carry a program that cannot occupy space by itself, such as a flexible disk, hard disk, CD-ROM, MO (magneto-optical disk), DVD-ROM, etc. is there. It is also possible to transmit from a computer installed with the program according to the present invention to another computer through a communication line. Moreover, it is also possible to form the fatigue analysis apparatus of the present invention by preinstalling or downloading the above program to a general-purpose terminal apparatus.

Test example 1
(Verification test of whether the use of the maximum Lyapunov index is effective)
Used by two Japanese adult subjects (subject A: male, 32 years old, height 178 cm, weight 69.2 kg, subject B: male, 34 years old, height 167 cm, weight 68 kg) in the driver's seat of a wooden chair or passenger car Fatigue in a state of sitting on a urethane sheet for 30 minutes was analyzed. In addition, the experiment was also performed when the robot was seated statically without being shaken, or when it was placed on a shaker and shaken. The excitation condition was a reproduction of the vibration of the floor during actual vehicle running with an average acceleration of 0.1G. A fingertip plethysmograph 20 was attached to each subject and a fingertip vessel pulse wave was collected, and a power value, LF / HF, and maximum Lyapunov index were determined, and a standard fatigue level, a corrected fatigue level, and an arousal level were determined. In Test Example 1, the standard fatigue level was calculated using an LF / HF inclusion power value in consideration of LF / HF time-series data of pulse rate. The slope of the power value, the slope of the corrected power value, and the slope of the maximum Lyapunov exponent were obtained with an overlap time of 162 seconds for 180 seconds. The results are shown in FIGS. Each of FIGS. 11 to 15 (a) also shows the sensory evaluation value of each subject. The sensory evaluation value uses the Borg scale. FIG. 11 shows the test results when the subject A is seated on the urethane sheet under static conditions, FIG. 12 shows the test results when the subject A is seated on the wooden chair under static conditions, and FIG. FIG. 14 shows the test results when subject A sits on the urethane sheet under vibration conditions, FIG. 14 shows the test results when subject A sits on the wooden chair under vibration conditions, and FIG. 15 shows static results when subject B sits on the wooden chair. The test results when seated under conditions are shown.

The “awareness fatigue” shown in each of FIGS. 11 to 15 is the output result of Test Example 1. This “subject fatigue level” is calculated by adding LF / L to the power value output by the power value slope integrating means 113 in the time zone before the time indicated by the alternate long and short dash line in each of FIGS. The cumulative fatigue level which is the cumulative sum of the standard fatigue levels calculated from the LF / HF calculated power value including HF is adopted, and in the time zone after the time indicated by the one-dot chain line, the second fatigue level calculating means 12 The correction power value calculation means 121 obtains a correction power value by multiplying the time series data value of the maximum Lyapunov exponent by the value of the time series data of the power value instead of the time series data of LF / HF, and calculates the correction power value. The corrected fatigue level is obtained from the value, and the cumulative fatigue level is adopted.

The time indicated by the alternate long and short dash line when the corrected fatigue level is adopted is based on the point when the tendency of sensory evaluation changes relatively, but the subjective fatigue level obtained in Test Example 1 is the sensory evaluation value. It can be seen that the curve is close to.

Each of FIGS. 11 to 15 (b) shows the fatigue curve (A) of the subjective fatigue degree shown in the figure of (a), and the cumulative sum of the standard fatigue degree calculated from only the power value in all time zones. Fatigue curve (B), the fatigue curve (C) of the cumulative sum of reference fatigue calculated from only the LF / HF power value in all time zones, the time before the time indicated by the one-dot chain line in FIG. Fatigue curve that combines the standard fatigue level obtained from the LF / HF power value in the time zone and the corrected fatigue level that is multiplied by the maximum Lyapunov exponent slope while adding LF / HF in the subsequent time zone ( D) is also shown. As a result, the fatigue curves (B) and (C) show different changes from the fatigue curve (A) of the subjective fatigue degree after the one-dot chain line time zone, and tend to deviate from the sensory evaluation value. It can be seen that the fatigue level closer to the sensory evaluation value can be obtained by adopting the corrected fatigue level including the maximum Lyapunov exponent at the time point. In the case of the fatigue curve (D), when compared with the fatigue curves (B) and (C), the sensory evaluation value is approached, but the time series data of LF / HF is included as the standard fatigue level For example, as shown in the fatigue curve (A) shown as the subjective fatigue level of Test Example 1, the time series data of the LF / HF is removed at the time when the time series change of the maximum Lyapunov slope is included. It turned out that it becomes a tendency close to evaluation.

Test example 2
(A test that automatically determines the maximum Lyapunov index inclusion timing)
In Test Example 2, as in Test Example 1, as the standard fatigue level, an LF / HF included power value obtained by adding LF / HF time-series data of the pulse rate to the power value was used. Further, in Test Example 2, the corrected fatigue that includes the maximum Lyapunov exponent using the determination means 13 including the arousal degree calculation means 131, the cumulative arousal degree output means 132, and the timing determination means 133 of the above embodiment. The timing to adopt the degree was automatically judged. That is, the timing determination unit 133 determines the sudden conversion point of the cumulative arousal level obtained by the cumulative arousal level output unit 132, and after that time, the corrected power value calculation unit 121 of the second fatigue level calculation unit 12 performs LF / Instead of the time series data of HF, the value of the time series data of the maximum Lyapunov exponent is multiplied by the value of the time series data of the power value to obtain the corrected power value, the corrected fatigue value is obtained from the corrected power value, and further accumulated The fatigue level was output. As in Test Example 1, the combination of the cumulative sum of the standard fatigue levels and the cumulative sum of the corrected fatigue levels was defined as the subjective fatigue level.

The “subject's subjective fatigue” shown in the diagrams (a) of FIGS. 16 to 20 is the output result of Test Example 2. Each of FIGS. 16 to 20 (b) is a diagram showing time-series changes in the “arousal level / time” and the “cumulative arousal level”. “Rounding degree / time” is an integral value every 18 seconds output by the maximum Lyapunov exponent slope integrating means 131c, and “cumulative arousing degree” is a cumulative sum of the integral values. In FIGS. 16 to 20A, the time indicated by the alternate long and short dash line is the time point when the timing determination means 133 determines that the corrected fatigue level is to be adopted. The timing determination unit 133 obtains the average inclination of the cumulative arousal level as the reference inclination for 4.45 minutes after the start of measurement for 4.45 minutes (the horizontal scale on the horizontal axis of each (b)), and thereafter The average inclination every 2.225 minutes (one scale on the horizontal axis in the figure of each (b)) was compared with the reference inclination, and the time when a change of 10% or more occurred was determined as the sudden conversion point. In each of FIGS. 16 to 20 (b), the time point indicated by the symbol “X” is the sudden conversion point of the slope, and the time point indicated by the symbol “Y” (the diagram of FIG. 16 to FIG. 20 (a)). Is the same as the point in time indicated by the alternate long and short dash line), the adoption timing of the corrected fatigue, that is, the maximum Lyapunov exponent. For example, referring to FIG. 16 (b), with 3.3 to 7.75 minutes (up to the second scale on the horizontal axis) as the reference inclination, the next 2.225 minutes with respect to this reference inclination, that is, 7. Compare the average slope from 75 minutes to 9.975 minutes. At this time, it can be judged when 9.975 minutes have elapsed that the average slope from 7.75 minutes to 9.975 minutes has changed by 10% or more with respect to the reference slope from 3.3 minutes to 7.75 minutes. . Therefore, the sudden conversion point X is set to a time point of 7.75 minutes, and the inclusion timing is when 9.975 minutes have elapsed.

FIG. 16 shows the test results when subject A sits on the urethane sheet under static conditions, FIG. 17 shows the test results when subject A sits on the wooden chair under static conditions, and FIG. FIG. 19 shows the test results when subject A sits on the urethane sheet under vibration conditions, FIG. 19 shows the test results when subject A sits on the wooden chair under vibration conditions, and FIG. 20 shows static results when subject B sits on the wooden chair. The test results when seated under conditions are shown.

As is apparent from this figure, it can be seen that a fatigue curve with a subjective fatigue level close to sensory evaluation can be obtained even when the timing of the corrected fatigue level is automatically determined by the timing determination means 133 from the slope of the cumulative arousal level. Therefore, when the configuration of Test Example 2 is adopted, for example, the degree of fatigue of the subject sitting on the seat can be automatically and accurately output.

In addition, each graph of (c) in FIGS. 16 to 20 shows the fatigue curve (A) of the subjective fatigue level shown in the graph of (a), and the reference fatigue level calculated from only the power value in all time zones. Cumulative sum fatigue curve (B), the fatigue curve (C) of the cumulative sum of reference fatigue calculated from only the LF / HF power value in all time zones, from the time indicated by the one-dot chain line in the diagram of (a) Fatigue combining the standard fatigue level obtained from the LF / HF power value in the previous time zone and the corrected fatigue level multiplied by the maximum Lyapunov exponent slope while adding LF / HF in the subsequent time zone Although the curve (D) is also shown, the time series data of the LF / HF at the time when the time series change of the maximum Lyapunov slope is included as in the fatigue curve (A) shown as the subjective fatigue level of Test Example 2 It ’s better to remove It was found to be in close trend in performance evaluation.

Test example 3
(A test in which the maximum Lyapunov index was counted from the “degree of arousal / time”)
In Test Example 3, as in Test Examples 1 and 2, as the standard fatigue level, an LF / HF included power value obtained by adding time-series data of LF / HF of the pulse rate to the power value was used. In Test Example 3, a predetermined threshold value is set for the “degree of arousal / time”, which is an integral value every 18 seconds output by the maximum Lyapunov exponent slope integration means 131c, and the LF / Instead of the time series data of HF, the corrected fatigue degree including the maximum Lyapunov exponent was adopted. The combination of the standard fatigue level and the corrected fatigue level is shown as the subjective fatigue level in FIGS. 21 to 25 (a).

21 to 25, each broken line shown in parallel with the horizontal axis is the threshold value. The threshold value refers to the change in “degree of arousal / time” from 4. 3 minutes after the start of measurement to 4.45 minutes (2 scales on the horizontal axis in each figure (b)). The measurer set a value slightly higher than the fluctuation range of the “arousal level / time” of the belt. 21 shows the test results when subject A sits on the urethane sheet under static conditions, FIG. 22 shows the test results when subject A sits on the wooden chair under static conditions, and FIG. FIG. 24 shows the test results when subject A sits on a urethane sheet under vibration conditions, FIG. 24 shows the test results when subject A sits on a wooden chair under vibration conditions, and FIG. 25 shows static results when subject B sits on a wooden chair. The test results when seated under conditions are shown. Each of FIGS. 21 to 25 (b) shows the fatigue curve (A) of the subjective fatigue degree shown in the figure (a), and the reference fatigue degree calculated from only the power value in all time zones. Cumulative sum fatigue curve (B), the fatigue curve (C) of the cumulative sum of reference fatigue calculated from only the LF / HF power value in all time zones, from the time indicated by the one-dot chain line in the diagram of (a) Fatigue combining the standard fatigue level obtained from the LF / HF power value in the previous time zone and the corrected fatigue level multiplied by the maximum Lyapunov exponent slope while adding LF / HF in the subsequent time zone The curve (D) is also shown. Each of FIG. 21 to FIG. 25 (c) is a diagram showing time-series changes in “degree of arousal / time” and “cumulative degree of arousal”.

As is clear from these figures, a predetermined time zone after the start of measurement, preferably a time zone of 3 to 12 minutes (4.45 minutes in this test example) in the range of 3 to 15 minutes after the start of measurement. It can be seen that a fatigue curve with a subjective fatigue level close to sensory evaluation can be obtained even if the threshold value is set with reference to the change in “arousal level / time”. As in Test Example 2, the configuration in which the maximum Lyapunov exponent is automatically calculated from the change in the slope of the cumulative arousal level is more computationally intensive than the manual setting of the threshold value based on the change in “Rounding level / time” in Test Example 3. In view of the fact that the value is large, depending on the computing ability of the fatigue analysis device, it is also possible to obtain the degree of subjective fatigue by the method of Test Example 3. However, as long as you set a ratio (10%, 20%, etc.) to be judged as a sudden conversion point of the slope of the cumulative arousal level, it automatically determines the timing of inclusion of the maximum Lyapunov index and outputs a subjective fatigue level close to sensory evaluation The technique of Test Example 2 is more preferable.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Fatigue analysis apparatus 10 Data receiving means 11 First fatigue degree calculating means 111 Power value calculating means 111a LF / HF time series data calculating means 112 Power value inclination calculating means 113 Power value inclination integrating means 12 Second fatigue degree calculating means 121 Correction Power value calculation means 122 Correction power value inclination calculation means 123 Correction power value inclination integration means 13 Determination means 131 Arousal degree calculation means 131a Maximum Lyapunov exponent calculation means 131b Maximum Lyapunov exponent inclination calculation means 131c Maximum Lyapunov exponent inclination integration means 132 Cumulative arousal degree Output means 133 Timing determination means 14 Cumulative fatigue degree calculation means 20 Fingertip plethysmograph

Claims

A fatigue analysis device that performs fatigue analysis using biosignal data of pulse waves collected by a biosignal measuring device,
The difference between the peak value on the upper limit side and the peak value on the lower limit side is calculated for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data, and this difference is used as the power value, and the time series of the power values is calculated. A first fatigue level calculating means for obtaining a standard fatigue level based on the data;
Using the maximum Lyapunov exponent determined from the biological signal data, the time-series data value of the power value used in the first fatigue level calculating means is corrected, and based on the obtained time-series data of the corrected power value. A second fatigue level calculating means for obtaining a corrected fatigue level,
A determination means for determining the presence or absence of compensation for fatigue due to sympathetic nerve activity;
In a time zone in which it is determined by the determination means that the fatigue due to sympathetic nerve activity has not been compensated, a cumulative sum of the reference fatigue levels obtained by the first fatigue level calculation means is obtained and output, and the sympathetic nerves are output. A cumulative fatigue level output means for obtaining and outputting a cumulative sum of corrected fatigue levels obtained by the second fatigue level calculation means in a time zone in which it is determined that the fatigue due to the activity is being compensated; Fatigue analysis device characterized by
The determination means obtains a maximum Lyapunov exponent from the vital sign data, and calculates a degree of arousal based on time series data of the maximum Lyapunov exponent, and arousal degree obtained by the arousal degree calculation means And a timing determination means for determining the timing of adopting the corrected fatigue level of the second fatigue level calculation means,
The cumulative fatigue level output unit is obtained by the first fatigue level calculation unit in a time zone before the timing determination unit determines that the corrected fatigue level of the second fatigue level calculation unit is adopted. The corrected fatigue obtained by the second fatigue degree calculating means is obtained in the time zone after obtaining and outputting the cumulative sum of the reference fatigue degrees and determining that the corrected fatigue degree of the second fatigue degree calculating means is adopted. The fatigue analysis apparatus according to claim 1, wherein the fatigue analysis apparatus is set to obtain and output a cumulative sum of degrees.
The first fatigue level calculating means calculates a difference between the peak value on the upper limit side and the peak value on the lower limit side for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data, and calculates the difference. A power value calculating means for calculating time series data of the power value as the power value;
A power value inclination calculating means for obtaining an inclination of the power value with respect to a time axis in a predetermined time range by performing slide calculation a predetermined number of times;
A power value slope integrating means for performing absolute value processing on a time series signal of power value slopes obtained by slide calculation by the power value slope calculating means, and calculating an integral value for each predetermined time range as the reference fatigue level; The fatigue analysis apparatus according to claim 2, wherein the fatigue analysis apparatus is provided.
Furthermore, it comprises LF / HF time series data calculating means for obtaining time series data of the power spectrum of LF / HF obtained by frequency analysis of time series data of heart rate or pulse rate collected by the biological signal measuring device,
The power value calculating means multiplies the value of the time series data of the power spectrum of the LF / HF and the value of the time series data of the power value by values at times corresponding to each other, and LF / HF 4. The fatigue analysis apparatus according to claim 3, wherein the fatigue analysis apparatus is set to calculate a time series data of the LF / HF inclusion power value as an HF inclusion power value.
The second fatigue level calculating means uses the value of the time series data of the maximum Lyapunov exponent used in the arousal level calculating means and the value of the time series data of the power value used in the first fatigue level calculating means. Correction power value calculating means for multiplying the values at the time corresponding to, and making it a correction power value, and calculating time series data of the correction power value;
Correction power value inclination calculating means for obtaining the inclination of the correction power value with respect to the time axis in a predetermined time range by performing slide calculation a predetermined number of times;
Corrected power value slope integrating means for performing absolute value processing on the time series signal of the slope of the corrected power value obtained by the slide calculation by the corrected power value calculating means, and calculating an integral value for each predetermined time range as the corrected fatigue level The fatigue analysis apparatus according to claim 1, comprising:
The second fatigue degree calculating means uses the value of the time series data of the maximum Lyapunov exponent used in the arousal degree calculating means and the value of the time series data of the LF / HF included power value used in the first fatigue degree calculating means. And a correction power value calculation means for calculating the time series data of the correction power value by multiplying the values at the time corresponding to each other as a correction power value,
Correction power value inclination calculating means for obtaining the inclination of the correction power value with respect to the time axis in a predetermined time range by performing slide calculation a predetermined number of times;
Corrected power value slope integrating means for performing absolute value processing on the time series signal of the slope of the corrected power value obtained by the slide calculation by the corrected power value calculating means, and calculating an integral value for each predetermined time range as the corrected fatigue level The fatigue analysis apparatus according to claim 4, comprising:
The arousal degree calculating means obtains the maximum Lyapunov exponent from the biological signal data, and calculates the maximum Lyapunov exponent time series data of the maximum Lyapunov exponent;
A maximum Lyapunov exponent slope calculating means for obtaining a slope of the maximum Lyapunov exponent with respect to a time axis in a predetermined time range by sliding a predetermined number of times;
Maximum Lyapunov exponent slope integration means for processing the absolute value of the time series signal of the slope of the maximum Lyapunov exponent obtained by the slide calculation by the maximum Lyapunov exponent slope calculation means and calculating the integral value for each predetermined time range as the arousal level And
3. The fatigue analysis apparatus according to claim 2, further comprising cumulative arousal output means for obtaining and outputting a cumulative sum of the arousal levels obtained by the maximum Lyapunov exponent slope integration means.
The timing determination unit determines that the sudden conversion point of the slope of the cumulative evoke degree output by the cumulative eviction level output unit is a timing at which the corrected fatigue level of the second fatigue level calculation unit is adopted. The fatigue analysis apparatus according to claim 2.
The timing determination means obtains the slope of the cumulative arousal degree in a predetermined time range from the start of measurement as a reference slope, and the slope of the cumulative arousal degree in a predetermined time range in the time zone after obtaining the reference slope is a reference slope. The fatigue analysis apparatus according to claim 7, wherein the sudden conversion point is determined when a change of at least 10% occurs.
A computer program installed in a fatigue analysis apparatus that performs fatigue analysis using biosignal data of a pulse wave collected by a biosignal measuring device,
The difference between the peak value on the upper limit side and the peak value on the lower limit side is calculated for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data. A first fatigue level calculating means for obtaining a standard fatigue level based on the data;
Using the maximum Lyapunov exponent determined from the biological signal data, the time-series data value of the power value used in the first fatigue level calculating means is corrected, and based on the obtained time-series data of the corrected power value. A second fatigue level calculating means for obtaining a corrected fatigue level,
A determination means for determining the presence or absence of compensation for fatigue due to sympathetic nerve activity;
In the time zone in which it is determined by the determination means that the fatigue due to sympathetic nerve activity has not been compensated, the cumulative sum of the reference fatigue levels obtained by the first fatigue level calculation means is obtained and output, and the sympathetic nerves are output. A cumulative fatigue level output means for obtaining and outputting a cumulative sum of corrected fatigue levels obtained by the second fatigue level calculation means in a time zone in which it is determined that the compensation for fatigue due to activity has been made; A computer program characterized by the above.
The determination means obtains a maximum Lyapunov exponent from the vital sign data, and calculates a degree of arousal based on time series data of the maximum Lyapunov exponent, and arousal degree obtained by the arousal degree calculation means And a timing determination means for determining the timing of adopting the corrected fatigue level of the second fatigue level calculation means,
The cumulative fatigue level output unit is obtained by the first fatigue level calculation unit in a time zone before the timing determination unit determines that the corrected fatigue level of the second fatigue level calculation unit is adopted. The corrected fatigue obtained by the second fatigue degree calculating means is obtained in the time zone after obtaining and outputting the cumulative sum of the reference fatigue degrees and determining that the corrected fatigue degree of the second fatigue degree calculating means is adopted. 11. The computer program according to claim 10, wherein the computer program is set so as to obtain and output a cumulative sum of degrees.
The first fatigue level calculating means calculates a difference between the peak value on the upper limit side and the peak value on the lower limit side for each predetermined time range from the peak value of each cycle of the original waveform of the biological signal data, and calculates the difference. A power value calculating means for calculating time series data of the power value as the power value;
A power value inclination calculating means for obtaining an inclination of the power value with respect to a time axis in a predetermined time range by performing slide calculation a predetermined number of times;
A power value slope integrating means for performing absolute value processing on a time series signal of power value slopes obtained by slide calculation by the power value slope calculating means, and calculating an integral value for each predetermined time range as the reference fatigue level; The computer program according to claim 11, comprising the computer program.
Furthermore, it comprises LF / HF time series data calculating means for obtaining time series data of the power spectrum of LF / HF obtained by frequency analysis of time series data of heart rate or pulse rate collected by the biological signal measuring device,
The power value calculating means multiplies the value of the time series data of the power spectrum of the LF / HF and the value of the time series data of the power value by values at times corresponding to each other, and LF / HF 13. The computer program according to claim 12, wherein the computer program is set so as to calculate time series data of the LF / HF included power value as an HF included power value.
The second fatigue level calculating means uses the value of the time series data of the maximum Lyapunov exponent used in the arousal level calculating means and the value of the time series data of the power value used in the first fatigue level calculating means. Correction power value calculating means for multiplying the values at the time corresponding to, and making it a correction power value, and calculating time series data of the correction power value;
Correction power value inclination calculating means for obtaining the inclination of the correction power value with respect to the time axis in a predetermined time range by performing slide calculation a predetermined number of times;
Corrected power value slope integrating means for performing absolute value processing on the time series signal of the slope of the corrected power value obtained by the slide calculation by the corrected power value calculating means, and calculating an integral value for each predetermined time range as the corrected fatigue level The computer program according to claim 10, comprising:
The second fatigue degree calculating means uses the value of the time series data of the maximum Lyapunov exponent used in the arousal degree calculating means and the value of the time series data of the LF / HF included power value used in the first fatigue degree calculating means. And a correction power value calculation means for calculating the time series data of the correction power value by multiplying the values at the time corresponding to each other as a correction power value,
Correction power value inclination calculating means for obtaining the inclination of the correction power value with respect to the time axis in a predetermined time range by performing slide calculation a predetermined number of times;
Corrected power value slope integrating means for performing absolute value processing on the time series signal of the slope of the corrected power value obtained by the slide calculation by the corrected power value calculating means, and calculating an integral value for each predetermined time range as the corrected fatigue level The computer program according to claim 13, comprising:
The arousal degree calculating means obtains the maximum Lyapunov exponent from the biological signal data, and calculates the maximum Lyapunov exponent time series data of the maximum Lyapunov exponent;
A maximum Lyapunov exponent slope calculating means for obtaining a slope of the maximum Lyapunov exponent with respect to a time axis in a predetermined time range by sliding a predetermined number of times;
Maximum Lyapunov exponent slope integration means for processing the absolute value of the time series signal of the slope of the maximum Lyapunov exponent obtained by the slide calculation by the maximum Lyapunov exponent slope calculation means and calculating the integral value for each predetermined time range as the arousal level And
12. The computer program according to claim 11, further comprising cumulative arousal level output means for obtaining and outputting the cumulative sum of the arousal levels obtained by the maximum Lyapunov exponent slope integrating means.
The timing determination unit determines that the sudden conversion point of the slope of the cumulative evoke degree output by the cumulative eviction level output unit is a timing at which the corrected fatigue level of the second fatigue level calculation unit is adopted. The computer program according to claim 11.
The timing determination means obtains the slope of the cumulative arousal degree in a predetermined time range from the start of measurement as a reference slope, and the slope of the cumulative arousal degree in a predetermined time range in the time zone after obtaining the reference slope is a reference slope. The computer program according to claim 16, wherein the sudden conversion point is determined when a change of at least 10% occurs.