JP6710357B1 - Exercise support system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a motion support system capable of appropriately transmitting a stimulus that is a trigger for exercising and recognizing a direction to be moved for each motion site with a small number of stimuli. SOLUTION: A stimulating unit attached to a user and having a plurality of stimulating means that sequentially operates, and one or more sensor units that detect a movement of a predetermined part of the user's body and output a detection signal, Based on the detection signal of the sensor unit, a feedback necessity determination unit that determines whether or not the stimulating unit needs to be operated, and a timing of operating the stimulating unit based on the detection signal of the sensor unit. When it is determined by the feedback timing calculation unit that calculates and the feedback necessity determination unit that the stimulating unit needs to operate, a feedback signal that operates the stimulating unit is generated at the timing calculated by the feedback timing calculating unit. And a feedback signal generation unit that generates the feedback signal. [Selection diagram]

Description

The present invention relates to a system for activating a motor function such as a human walking function, and particularly to a motion support system capable of activating a human motor function by feeding back a human motion to a sensory function.

Conventionally, many robotics technologies have been developed for the purpose of assisting walking, and various manufacturers have marketed their own assisting devices for walking. In the case of walking assist using this robotics, the motion is sensed and the motion is physically assisted, but with the current technology, the physical movement including position, range of motion, balance, reaction, etc. It is a difficult situation to cover even the differences. Therefore, the assistance desired by the user has not been achieved, and there is a limit to the improvement of walking only with robotics technology.

However, attention is focused only on the physical walking assist of robotics, and it is not always sufficient to consider what should be assisted for a person with walking difficulty. For example, there is no discussion about what pedestrians are having trouble with, what they want to do, and what kind of technology is needed for that.

The above problem is considered to be because the walking assist function using the robotics technology is not linked with the walking function of humans. There are individual differences in the walking function of humans, and it also differs depending on the age, the presence or absence of injury, or the site of injury.

Based on the fact that this human walking function is utilized to improve it so that it can be activated, that is, a target motion, and if there is a demand to further raise the foot, for example, If the robotics technology is applied only to that part, the above problems would be considerably improved.

As a technique related to walking assist for the purpose of activating the walking function of humans, for example, in Patent Document 1, a sensor unit that detects a movement rhythm of a pedestrian and a measurement value of the movement rhythm detected by the sensor unit are provided. A recording unit that records, a target setting unit that sets a target value related to a pedestrian's movement rhythm, a timing generation unit that generates a timing signal based on a difference between the measured value and the target value, and a timing generated by the timing generation unit. A walking assistance system including a stimulus generator that generates a rhythmic stimulus that can be recognized by a pedestrian based on a signal is described. The technique of Patent Document 1 is intended to improve walking motion by causing a pedestrian to walk at a rhythm close to a target value by stimulation such as sound, light, and electricity.

Further, Patent Document 2 includes a biaxial acceleration sensor that detects acceleration in the traveling direction and vertical acceleration of the affected leg and a uniaxial angular velocity sensor that detects angular velocity around the knee axis, and visually observes the affected leg of each patient. There is described a walking assist device that highly accurately estimates the leaving timing of various foot drop patients with a neural network that learns the leaving timing as a teacher signal, and generates electrical stimulation at a good timing.

JP, 2006-102156, A JP, 2010-227236, A

However, in the above-described conventional technique, the stimulation signal is output in a single shot at a predetermined timing. For this reason, although the user can move the part that senses the stimulus, it is difficult for the user to know whether or not the direction was correct.

In addition, during exercise, the muscles of the body perform complicated movements while interlocking with each other. If the stimulation is given to each site each time the timing arrives, if the number of sites to be stimulated increases, the number of stimulations becomes too large and the effect decreases. Therefore, it becomes a problem how effectively a desired operation can be performed with a small number of stimuli.

Further, the above-mentioned conventional technique generates a stimulus to a walking user, but depending on the degree of the walking disorder, it is often difficult to shift from a stopped state to a working state at the start of walking. ..

The present invention has been made to address the above problems, and provides an exercise support system capable of appropriately transmitting a stimulus that triggers the start of exercise and recognizing a direction to move for each operation site with a small number of stimuli. Aim to

In order to achieve the above-mentioned object, a luck support system according to the present disclosure is an exercise support system that stimulates a user with exercise.
A memory for storing a first parameter regarding the operation timing of each body part linked to the movement of the reference part of the body obtained by machine learning and a second parameter regarding the amount of movement of each body part obtained by machine learning Department,
A stimulating unit that is attached to the user and has a plurality of stimulating means that operate in sequence,
One or more sensor units that detect a movement of a predetermined part of the user's body and output a detection signal,
A feedback necessity determination unit that determines whether or not the stimulating unit needs to be operated based on the detection signal of the sensor unit and the second parameter ,
A feedback timing calculation unit that calculates a timing for operating the stimulating means based on the first parameter and the detection signal;
When the feedback necessity determination unit determines that the operation of the stimulating unit is necessary, a feedback signal generating unit that generates a feedback signal for operating the stimulating unit at the timing calculated by the feedback timing calculating unit,
It is characterized by having.

In the present disclosure, the feedback necessity determination unit and the feedback timing calculation unit are independently executed. The feedback necessity determination unit operates the stimulation unit at the timing calculated by the feedback timing calculation unit only when it is determined that the operation of the stimulation unit is necessary. Thereby, the user can accurately know whether or not the original operation is performed.

Preferably, in the stimulating section, a plurality of stimulating means are arranged linearly or in a plane, and the feedback signal generating section starts operation from the stimulating means at a predetermined position and operates adjacent stimulating means in order. It is preferable to generate the feedback signal. More preferably, as the operation of the stimulation means, the direction of stimulation and the number of stimulations may be variable.

As a result, the user can be given a feeling of being stroked, and the user can know the original movement direction and the degree of deviation from the original movement.

In particular, the feedback determination unit may determine whether or not the stimulating means needs to be operated based on the time interval of the detection signal output from the sensor. Accordingly, for example, it is determined whether or not the correction is necessary based on the cycle of the user's exercise, the time difference between the operation timings of the user's body parts by the plurality of sensors, and the like.

The stimulating unit attached to the body of the user may include a unit that stores the amount of motion (stimulation amount) of the stimulating unit, and the amount of stimulation may be set externally.

Further, the exercise support system according to the present disclosure is characterized in that the speed at which each stimulating unit operates continuously is determined based on the time interval of the detection signal output from the sensor. As a result, it becomes possible to effectively correct the motion by operating the stimulating unit according to the operation speed of the user.

In the present invention, the feedback necessity determination unit and the feedback timing calculation unit are executed independently, and when the feedback necessity determination unit determines that the operation of the stimulating means is necessary, the timing calculated by the feedback timing calculation unit. Since the stimulating means is operated by, the stimulus that triggers the start of exercise can be appropriately transmitted, and the stimulus can be given only when improvement of the motion is necessary.

Further, by operating each of the plurality of stimulating means in order in the predetermined direction with a time difference, the user can easily know the direction and the amount to be moved with respect to the target site.

FIG. 1 is a system configuration diagram according to an embodiment of the present invention. It is a block diagram of the feedback apparatus of FIG. FIG. 3 is a layout view of stimulation means of the feedback device of FIG. 2. It is operation|movement explanatory drawing of a stimulating means when a leg adduction during walking. It is explanatory drawing of the operation example of a walking stage and a stimulation means. (Part 1) It is explanatory drawing of the operation example of a walking stage and a stimulation means. (Part 2) It is explanatory drawing of the operation example of a walking stage and a stimulation means. (Part 3) 6 is a flowchart showing a processing procedure of a feedback determination function 32 of the arithmetic processing device 3. It is explanatory drawing of the example of use data of a feedback judgment function, and is an example of the walking timing data (data 1) and walking model data (data 2) which have been machine-learned. It is explanatory drawing of the usage data example of a feedback judgment function, and is an example of sensor data (data 3). It is explanatory drawing of the usage data example of a feedback judgment function, and is an example of the timing analysis data (data 4) by a neural network process. It is explanatory drawing of the example of use data of a feedback judgment function, and is an example of the gait analysis data (data 5) by a neural network process. It is explanatory drawing of the usage data example of a feedback judgment function, and is an example of the data which shows a feedback part. It is explanatory drawing of the relationship between the timing calculated by the neural network process (timing analysis) and the output timing of the stimulation signal by the calculation result of a neural network process (walking analysis).

Hereinafter, the exercise support system according to the present embodiment will be described by taking a walking exercise as an example.
As shown in FIG. 1, the exercise support system 1 basically includes a sensor device 2, an arithmetic processing device 3, and a feedback device 4. Further, the arithmetic processing unit 3 has an arithmetic unit that executes various functions and a storage unit that stores data, and a server that collects exercise data via the communication network 6 and calculates the parameters of the exercise model. Connect with device 5.

The arithmetic processing unit 3 collects data such as time, acceleration, gyro, and geomagnetism from the data transmitting/receiving function 33 that communicates with the server unit 5 and the sensor unit 2 to calculate the three-dimensional acceleration of a predetermined part of the user and the rotational movement of each axis. A feedback determination function that determines whether or not feedback is necessary using the data acquisition function 31 to acquire, parameters (feedback determination parameters) downloaded from the server device 5, and outputs a feedback command to the feedback device 4 when feedback is required. 32 is provided. The arithmetic processing device 3 may be a dedicated computer device, but may be a general-purpose mobile terminal such as a smartphone, for example.

FIG. 2 is a block diagram of the feedback device 4. As shown in this figure, the feedback device 4 includes a stimulation unit 41 having a plurality of stimulation units a1 to a5. The stimulus output from the stimulating means a1 to a5 may be electrical or vibration stimulus, or pressure stimulus with variably controlled air pressure. In response to a command (feedback command) output from the feedback determination function 32, the stimulating means a1 to a5 operate in a predetermined order to give a stimulus through the user's skin sensation. It should be noted that the stimulating means a1 to a5 of the stimulating section 41 may be ones arranged two-dimensionally as shown in FIG.

The amount of stimulation to the skin can be adjusted by changing the amount of movement of the stimulation means a1 to a5 of the stimulation unit 41. This adjustment can be realized, for example, by selecting each of the stimulating means a1 to a5 with the switches A to D of the operation unit 42 and changing the operation amount with the volume. The operation amount and the timing adjustment amount for each stimulating means set by the operation unit 42 are sent to the stimulating unit 41 and stored therein. It should be noted that any number of stimulation means can be used depending on the user and the stimulation site.

Next, the operation of the exercise support system 1 will be described.
[Operation overview]
The exercise support system 1 downloads parameter data to be compared in advance from the server device 5 and stores it in the storage unit of the arithmetic processing device 3. Then, the walking data is acquired from the sensor device 2 attached to the user's knees and toes.
The feedback determination function 32 of the arithmetic processing device 3 compares the parameter data with the walking data of the user, and when it is determined that feedback is necessary, the feedback determination function 32 of the feedback device 4 mounted at a predetermined position such as the user's knee or elbow. For example, linearly provided stimulating means a1 to a5 are operated in the stimulating unit 41 to give a stimulus through the user's skin sensation.

As an example of how to give a stimulus, when the knee does not go up (as a result, the stride becomes narrow), the stimulating section 41 is attached to the upper part of the knee. Then, in synchronization with the walking timing, the stimulus of tapping the knee is first applied to concentrate the consciousness on the stimulated site.

When the legs adduction during walking, as shown in FIG. 4, as the knee lifting stimulus, the stimulating means a6 to a6 in the knee-surrounding direction are sequentially operated at the timing of sequentially operating the stimulating means a1 to a5 from below the knee to above. The a10 is operated at predetermined time intervals to sequentially give stimulation from the inside of the knee to the outside. As a result, the user can feel the stimulus of being stroked from the inside to the outside of the knee and can be aware of the rotation motion of the leg from the inside to the outside. The trigger that starts the series of stimulating motions can be set to a certain timing during the walking motion, for example, a timing at which the ground kicks out.

It is effective to apply stimulation below the knees from above and above the insteps from the toes to the ankles in conjunction with stimulation of the upper knee. At this time, the operation unit 42 is adjusted in advance so as to strongly stimulate the place to be most conscious of for each user. In addition, a walking stage and examples of stimulating motions are shown in FIGS.

The feedback signal generation unit starts the operation from the stimulating means at a predetermined position, generates a feedback signal so as to sequentially operate the adjacent stimulating means, and when the stimulating means is sequentially moved, the operation of the stimulating means at the predetermined position. It is also possible to perform a so-called advance notice operation in which only the stimulating means at the predetermined position is activated once or several times before the start of the. At this time, it is effective to make the stimulation amount larger than the original stimulation amount. As a result, the user can direct his/her attention to the part, and can sense even a small amount of stimulation at the subsequent original timing.

[Processing contents of the feedback judgment function]
FIG. 8 shows a processing procedure of the feedback judgment function which is the most important in this embodiment. 9A to 9E show examples of usage data of the feedback determination function. 9A to 9E, the rows (vertical direction) show items to be measured by the sensor device, and the columns (horizontal direction) show examples of input and output data in each process (function 1 to function 6). The data structure shown in FIGS. 9A to 9E may be configured by being divided into individual parts, or may be configured by one table, for example.

Note that the function numbers (functions 1 to 6) and data numbers (data 1 to 5) in the usage data examples of the feedback determination function shown in FIGS. 9A to 9E are described in the feedback determination function flowchart shown in FIG. It corresponds to the function number and data number.

Hereinafter, the operation of the exercise support system according to the present embodiment will be described in detail with reference to the processing procedure of FIG. 8 and the data configuration diagram of FIG. 9 (FIGS. 9A to 9E).

First, the parameter data that has been machine-learned for the walking timing and the walking model is acquired in advance, and is stored in the storage unit of the arithmetic processing device 3 as a feedback determination parameter. (S101, S201). It should be noted that these parameters are generated by the parameter generation function of the server device 5 and downloaded to the arithmetic processing device 3 of each user.

FIG. 9A shows an example of gait timing parameter data (data 1: gait timing data) and gait model parameter data (data 2: gait model data). The walking timing indicates the operation timing of each part linked to the movement of the reference part, and can be expressed in an arbitrary unit such as time or a ratio to the entire cycle. The walking model indicates the amount of movement of each part. It is preferable that the parameter data can be adjusted for each user on the arithmetic processing device 3 side.

One of the features of this embodiment is that the operation timing (data 1) and the operation amount (data 2) of each part are managed separately.

The processing of steps S301 to S306 in FIG. 8 is repeatedly executed at a cycle sufficiently faster than the operation cycle of the reference part.

First, in step S301, sensor data of user walking is acquired from the sensor device 2 and stored in the storage unit of the arithmetic processing device 3. The sensor data includes, but is not limited to, time, acceleration sensor value, gyro sensor value, geomagnetic sensor value, and the like. These data are acquired at a constant frequency (Hz). FIG. 9B is an example of sensor data. The walking timing data and the walking model data of the user acquired by the sensor device 2 are stored for each data item.

Then, as neural network processing, feedback timing data is generated (S302). In step S302, the neural network post-evaluation data, that is, the feedback timing data having the highest effect probability is used.

Further, in parallel with this, it is determined whether or not feedback is necessary (S303). In this processing, the data after evaluation of the neural network is used, and whether or not to feed back is calculated as a judgment value with a probability of 0% to 100%.

FIG. 9C is a walking timing value for each part calculated by the timing analysis process. With reference to a predetermined reference position, for example, the landing time point of the sole of the right foot, the optimum timing value for the user is stored from there. FIG. 9D shows the difference between the walking model data (data 2) and the walking model data (data 3) of the user, which is calculated by the walking analysis process, and the determination value (% value) of the necessity of feedback. In the example of FIG. 9D, the larger the% value, the more the feedback is required.

Based on the data 5, the calculation unit of the calculation processing device 3 determines whether or not the threshold value of the probability determined in advance for each part is exceeded (S304), and if it exceeds, a feedback process is executed. (S305). For example, when the threshold value is set to 70% in step S304, a circle is attached to the feedback part in the data 5 of FIG. 9D where the feedback necessity judgment result is 70% or more, that is, the walking posture correction feedback part of FIG. 9E. A feedback signal (operation command) is output to the stimulating means at the selected site. This feedback signal is generated so that the stimulating means operates sequentially from a predetermined position based on the walking timing of each part of the data 4 and the value difference of the data 5.

Through the above processing, the stimulating means is provided only for the part requiring feedback. In the above process, steps S303 to S304 configure a feedback necessity determination unit, and step S302 configures a feedback timing calculation unit. Moreover, step S305 comprises a feedback signal generation part.

[Characteristics of processing of feedback judgment function]
The feedback judgment function described above has the following two major characteristics.
(First feature)
The following two neural network processes are executed for each stimulation target site (upper knee, lower knee, instep, sole) in this example.
1. Feedback necessity judgment by neural network processing (walk analysis) <Function 5>
2. Feedback timing calculation by neural network processing (timing analysis) <Function 4>

(Second feature)
If there is a stimulation target site determined to require feedback by the neural network processing (walking analysis), a feedback signal is generated for the stimulation target site using the outputs of the two neural network processes. <Function 6>

More specifically, the stimulation means is operated based on the stimulation direction and the number of stimulations calculated by the neural network processing (walking analysis) at the timing calculated by the neural network processing (timing analysis).

FIG. 10 shows the relationship between the timing calculated by the neural network processing (timing analysis) and the output timing of the stimulation signal based on the calculation result of the neural network processing (walking analysis).

In FIG. 10, the timing difference T is a fixed value (positive value, negative value, 0) for each stimulation site and each user. The strength H of the stimulus can be set from the outside via an input device (for example, the operation unit 42). However, it may be variable depending on the magnitude of the difference between the measured value and the original value (parameter data).

Although (1) to (5) of FIG. 10 show identification numbers of a plurality of stimulating means, the order and number in this figure are examples. For example, if you want to move in the opposite direction, you may go from (5) to (1), or if you want to increase the amount of movement, for example, when you raise your knee, raise it further. When urging the user to urge, the stimulus may be given again by returning to (1) after (1) to (5). Alternatively, the movement can start from (3) and spread in both directions (2)(1) and (4)(5). Further, the stimulating means is not limited to one dimension, and it is of course possible to arrange it in two dimensions.

Conventionally, only neural network processing (timing analysis) was performed, but in the present embodiment, the feedback determination function having the first and second characteristics described above causes stimulation in synchronization with the timing signal as shown in FIG. Generating the operating signal of the means.

Synchronous with the timing calculated by the neural network processing (timing analysis) for each stimulation target site, the stimulating means is activated based on the stimulation direction and the amount of stimulation calculated by the neural network processing (walking analysis). Can properly convey the stimulus that becomes.

Also, based on the calculation results of a plurality of neural network processes (walking analysis), a plurality of stimulating means are operated with a time lag in a predetermined direction, so that the user can easily know the direction and amount to move with respect to the target part. be able to.

Neural network processing (walking analysis) determines whether or not feedback is required for each stimulation site, and the stimulation means is not activated for areas that do not require feedback, so the number of stimulations is reduced without affecting the walking function. can do.

The present invention can be used for activation of the walking function, but is not limited to this, and can be widely applied to general human motion. In addition, it can be used not only for people with walking difficulties but also for exercise training for healthy people and the like.

For example, when the stimulating means is attached to the arm, knee or torso and the transition from the golf backswing to the downswing is performed, the timing of the turning back and the direction of the downswing are set to the timing and the operation direction input as the teacher data. It is also possible to feed back how much the difference is.

1 Exercise Support System 2 Sensor Device 3 Arithmetic Processing Device 4 Feedback Device 5 Server Device 6 Communication Network 31 Data Acquisition Function 32 Feedback Judgment Function 33 Data Transmission/Reception Function 41 Stimulation Unit 42 Operation Unit a1-a10 Stimulation Means

Claims (4)

An exercise support system for stimulating a user with exercise, comprising:
A memory for storing a first parameter regarding the operation timing of each body part linked to the movement of the reference part of the body obtained by machine learning and a second parameter regarding the amount of movement of each body part obtained by machine learning Department,
A stimulating unit that is attached to the user and has a plurality of stimulating means that operate in sequence,
One or more sensor units that detect a movement of a predetermined part of the user's body and output a detection signal,
A feedback necessity determination unit that determines whether or not the stimulating unit needs to be operated based on the detection signal of the sensor unit and the second parameter ,
A feedback timing calculation unit that calculates a timing for operating the stimulating means based on the first parameter and the detection signal;
When the feedback necessity determination unit determines that the operation of the stimulating unit is necessary, a feedback signal generating unit that generates a feedback signal for operating the stimulating unit at the timing calculated by the feedback timing calculating unit,
An exercise support system comprising: In the stimulator, a plurality of stimulators are arranged linearly or in a plane, and the feedback signal generator starts the operation from the stimulator at a predetermined position and sequentially operates the adjacent stimulators. The exercise support system according to claim 1, wherein: The exercise support system according to claim 1 or 2 , wherein the stimulation unit includes a unit that stores the amount of movement of the stimulation unit, and the amount of movement can be set from the outside. The movement amount stored in the stimulation unit can be adjusted for each stimulation unit,
The exercise support system according to claim 3 , further comprising: an operation unit having a unit that selects the stimulating unit and a unit that adjusts an operation amount of the selected stimulating unit.