JP2019203670A - Air conditioner and method - Google Patents

Abstract

To provide an air conditioner which can make an indoor temperature reach a set temperature in a shorter time, and a method.SOLUTION: A VAV controller 1 comprises: a requirement air quantity calculation part 14 for calculating a requirement air quantity for making an indoor temperature of an indoor area 6 reach a set temperature; an air quantity control part 16 for creating a control signal indicating a duct opening according to a deviation between the requirement air quantity and a measurement air quantity of intake air from a VAV unit 4; and an NN part 18 for learning a relationship of the requirement air quantity and the measurement air quantity by using a neural network, and constructing the learnt neural network for estimating a correction amount of the requirement air quantity. The NN part 18 estimates the correction amount by using the learnt neural network, and the air quantity control part 16 creates the control signal for indicating the duct opening for adjusting the air quantity of the intake air from the VAV unit 4 according to the deviation between the corrected requirement air quantity which is corrected according to the estimated correction amount and the measurement air quantity.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an air conditioning control apparatus and method, and more particularly to an air conditioning control technology using a neural network.

  The load reset control of the conventional variable air volume (VAV) system automatically adjusts the supply temperature setting value of the air conditioner according to the load state of the VAV unit so that the room temperature set by the user is obtained. It is control (refer patent document 1).

  For example, when a person concentrates in a conference room and the room temperature rises and the user feels hot and the room temperature setting is lowered, a difference occurs between the room temperature and the set temperature. In such a case, the VAV controller opens the damper of the VAV unit and controls the set air volume so that the room temperature approaches the set temperature. When the damper opening of the VAV unit is fully opened, the supply air temperature setting value is controlled so as to lower the supply air temperature of the air conditioner. Through these operations, the VAV system inputs only a necessary amount of heat into the room and realizes energy saving.

  In a new building, determine the maximum airflow of the air conditioner, the size of the duct, the minimum and maximum airflow of the VAV unit at the time of instrumentation design, and confirm the performance such as the maximum airflow of the air conditioner and the maximum airflow of the VAV unit during trial operation To adjust the parameters of the VAV controller. However, if there is a problem with the duct work, there is a problem of duct pressure loss, and even if the air volume of the air conditioner is set to the maximum air volume, the maximum air volume of the VAV unit may not be achieved.

  In such a case, if the relationship between the air volume of the air conditioner and the air volume of the VAV unit is closely investigated and the parameters of the VAV controller are finely adjusted, the response performance of the indoor temperature control can be executed as originally designed. it can.

Japanese Patent No. 6280456

  However, it takes an enormous amount of time to investigate the air conditioners and VAV units as described above and finely adjust the parameters of the VAV controller, so that it is difficult to carry out the adjustment at the site. In addition, the adjustment time tends to be shorter in recent sites, and the current situation is that no response can be made. As a result, there is a problem that even if the set temperature is changed, it takes more time than expected until the room temperature reaches the set temperature.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide an air conditioning control device and method capable of causing the room temperature to reach the set temperature in a shorter time.

  In order to solve the above-described problem, an air conditioning control device according to the present invention is an air conditioning control device that controls a variable air volume unit provided for each controlled area, and the indoor temperature of the controlled area is set to a set temperature. A required air volume calculation unit configured to calculate a required air volume, which is an air volume required for supplying air from the variable air volume unit, and measurement of the required air volume and supply air from the variable air volume unit. An air volume control unit configured to generate a control signal indicating a duct opening for adjusting the air volume of the supply air from the variable air volume unit according to a deviation from the air volume, the required air volume, and the measured air volume A neural network unit that constructs a learned neural network that learns the relationship between the The network unit estimates the correction amount using the learned neural network, and the air volume control unit responds to a deviation between the correction required air volume corrected according to the estimated correction amount and the measured air volume. Then, a control signal indicating a duct opening for adjusting the air volume of the supply air from the variable air volume unit is generated.

  Further, in the air conditioning control device according to the present invention, the air conditioning control device further includes a determination unit configured to determine whether or not there is a factor that causes variation in the air volume of the supply air from the variable air volume unit, and the neural network unit includes When the determination unit determines that the factor is present, the relationship between the required air volume and the measured air volume may be learned using a neural network.

  In the air conditioning control device according to the present invention, the air volume control unit may perform variable speed floating control according to the deviation to generate the control signal.

  The air conditioning control device according to the present invention may further include a first deviation calculating unit configured to calculate the deviation between the required air volume and the measured air volume.

  The air conditioning control apparatus according to the present invention may further include a second deviation calculating unit configured to calculate the deviation between the correction required air volume and the measured air volume.

  Further, in the air conditioning control device according to the present invention, the required air volume calculation unit calculates the required air volume by PID control using the set temperature as a target value, the indoor temperature as an input value, and the required air volume as an operation amount. Also good.

  The air conditioning control method according to the present invention is an air conditioning control method for controlling a variable air volume unit provided for each controlled area, and the indoor temperature of the controlled area reaches a set temperature. A required air volume calculating step for calculating a required air volume required for the air volume of the air supplied from the air volume unit, and a deviation between the required air volume and the measured air volume of the air supplied from the variable air volume unit, from the variable air volume unit. An air flow control step for generating a control signal indicating a duct opening for adjusting the air flow rate of the air supply; a determination step for determining the presence or absence of a factor that causes a variation in the air flow rate of the air supply from the variable air flow unit; When it is determined in the determination step that the factor is present, the relationship between the required air volume and the measured air volume of the supply air from the variable air volume unit is learned using a neural network. And learning steps, characterized in that it comprises a storage step of storing information about the trained neural network constructed in the learning step in a storage unit.

  The air conditioning control method according to the present invention is an air conditioning control method for controlling a variable air volume unit provided for each controlled area, and the indoor temperature of the controlled area reaches a set temperature. A required air volume calculating step for calculating a required air volume required for the air volume of the air supplied from the air volume unit, and an estimating step for estimating the correction amount of the required air volume using a learned neural network stored in the storage unit And the air volume of the supply air from the variable air volume unit according to the deviation between the correction required air volume corrected according to the correction amount estimated in the estimation step and the measured air volume of the air supply from the variable air volume unit And an air volume control step for generating a control signal indicating a duct opening degree for adjusting the air flow.

  According to the present invention, the relationship between the calculated required air volume and the actual measured air volume is learned using a neural network, the correction amount of the required air volume is estimated using the learned neural network, and the correction is performed based on the correction amount. Since the control signal indicating the damper opening degree of the VAV unit is generated using the corrected required air volume, the room temperature can reach the set temperature in a shorter time.

FIG. 1 is a block diagram showing a configuration of an air conditioning system including a VAV controller according to an embodiment of the present invention. FIG. 2 is a block diagram showing a functional configuration of the VAV controller according to the embodiment of the present invention. FIG. 3 is a diagram for explaining PID control according to the embodiment of the present invention. FIG. 4 is a diagram for explaining variable speed floating control according to the embodiment of the present invention. FIG. 5 is a block diagram showing a hardware configuration of the VAV controller according to the embodiment of the present invention. FIG. 6 is a flowchart for explaining the operation at the time of adjustment of the VAV controller according to the embodiment of the present invention. FIG. 7 is a flowchart for explaining the required air volume calculation processing according to the embodiment of the present invention. FIG. 8 is a flowchart illustrating the learning process according to the embodiment of the present invention. FIG. 9 is a flowchart illustrating the air volume control process according to the embodiment of the present invention. FIG. 10 is a flowchart for explaining the operation during operation of the VAV controller according to the embodiment of the present invention. FIG. 11 is a flowchart for explaining the estimation processing according to the embodiment of the present invention. FIG. 12 is a flowchart illustrating the air volume control processing according to the embodiment of the present invention.

  Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to FIGS. Constituent elements common to the drawings are given the same reference numerals. Hereinafter, the case where the “air conditioning control device” according to the present invention is a “VAV controller” will be described.

[Embodiment]
First, a configuration of an air conditioning system including a VAV controller 1 according to an embodiment of the present invention will be described with reference to FIG.
The air conditioning system includes a VAV controller 1, an air conditioning controller 2, an air conditioner 3, a VAV unit (variable air volume unit) 4, a duct 5, a temperature sensor 7, a wind speed sensor 8, and an outlet 9.

  The VAV controller 1 according to the present embodiment adjusts the amount of air supplied from the VAV unit 4 installed in each indoor area (controlled area) 6 according to the air conditioning load, and the indoor temperature in the indoor area 6 is adjusted. Control to reach set temperature. The VAV controller 1 adjusts the amount of air supplied from the VAV unit 4 by adjusting the opening degree of the damper 4-1 provided near the outlet 9 of the VAV unit 4.

  The air conditioning controller 2 manages and monitors the air conditioner 3 and the entire air conditioning system. The air conditioner 3 supplies conditioned air to the VAV unit 4. The VAV unit 4 adjusts the amount of air supplied to the indoor area 6 according to the opening degree of the damper 4-1. The VAV unit 4 includes an adjustment mechanism (not shown) that adjusts the opening degree of the damper 4-1, a sensor (not shown) that detects the damper opening degree, and the like. The VAV unit 4 adjusts the opening degree of the damper 4-1 based on the control signal sent from the VAV controller 1.

  The duct 5 supplies the air supply from the air conditioner 3 to each indoor area 6. The temperature sensor 7 detects the indoor temperature of the indoor area 6. The wind speed sensor 8 detects the wind speed of the conditioned air from the VAV unit 4.

[Function block of VAV controller]
FIG. 2 is a block diagram showing a functional configuration of the VAV controller 1 according to the present embodiment.
The VAV controller 1 includes a set temperature acquisition unit 10, an indoor temperature acquisition unit 11, a measured air volume acquisition unit 12, an actual maximum air volume acquisition unit 13, a required air volume calculation unit 14, and an air volume deviation calculation unit (first and second deviation calculation units). 15, an air volume control unit 16, a duct loss determination unit (determination unit) 17, a neural network (NN) unit 18, an output unit 19, and a storage unit 20.

  The set temperature acquisition unit 10 acquires a set temperature of the indoor area 6 input by the user or a preset temperature set in advance. The acquired set temperature is input to the required air volume calculation unit 14.

  The room temperature acquisition unit 11 acquires the room temperature measured by the temperature sensor 7. The acquired room temperature is input to the required air volume calculation unit 14.

  The measured air volume acquisition unit 12 acquires a measured value of the air volume of the supply air flowing through the VAV unit 4 installed in the indoor area 6 from the wind speed sensor 8 provided in the VAV unit 4 or the like. More specifically, the measured air volume acquisition unit 12 calculates the wind speed by converting the voltage output from the wind speed sensor 8, and calculates the air volume by multiplying it by the duct area. The measured air volume acquired by the measured air volume acquiring unit 12 is input to the air volume deviation calculating unit 15.

  The actual maximum air volume acquisition unit 13 is the measured air volume acquisition unit 12 when the supply air supplied from the air conditioner 3 to the plurality of indoor areas 6 is the maximum air volume and the damper 4-1 of the VAV unit 4 is fully opened. The measured air volume acquired by is acquired as the actual maximum air volume. The acquired actual maximum air volume is stored in the storage unit 20.

  The required air volume calculation unit 14 calculates a required air volume that is an air volume required for supplying air from the VAV unit 4 so that the room temperature in the indoor area 6 reaches the set temperature. More specifically, the required air volume calculation unit 14 calculates the required air volume based on the deviation between the room temperature and the set temperature. The calculated required air volume value is input to the air volume deviation calculating section 15 and the duct loss determining section 17. For example, the required air volume calculation unit 14 performs well-known PID control, and for each indoor area 6 to be controlled, the target value is the set temperature of the indoor area 6, the input indoor temperature is measured, and the operation amount is the required air volume. PID calculation is performed.

  For example, as shown in FIG. 3, the required air volume is determined by proportional operation according to the relationship between the operation amount (required air volume) and the room temperature in each operation mode during heating operation, air blowing operation, and cooling operation. Further, in the integration operation, the required air volume corresponding to a value obtained by dividing the integrated value by the integrated time is calculated as an integrated value by adding the deviation between the set temperature and the room temperature at regular intervals. Further, in the differential operation, the required air volume is corrected by outputting an output proportional to the differential of the deviation.

  Further, as shown in FIG. 3, the required air volume calculation unit 14 performs PID control with the minimum air volume and the actual maximum air volume of the VAV unit 4 as the minimum value and the maximum value of the output value of the required air volume, respectively. Note that the minimum air volume is a value set in advance by the manufacturer's shipping inspection of the VAV unit 4, and the actual maximum air volume is a value acquired by the above-described actual maximum air volume acquiring unit 13 and stored in the storage unit 20. It is.

  The air volume deviation calculating unit 15 calculates the air volume deviation based on the required air volume or a corrected required air volume, a measured air volume, and an actual maximum air volume that will be described later. The required air volume is the air volume calculated by the required air volume calculator 14. The corrected required air volume is a value of the required air volume that takes into account the variation in the air volume of the air supplied from the VAV unit 4 estimated by the NN unit 18 described later. The measured air volume is the actual air volume flowing from the VAV unit 4 acquired by the measured air volume acquisition unit 12. The actual maximum air volume is an air volume acquired by the actual maximum air volume acquisition unit 13 and stored in the storage unit 20.

When the air volume deviation calculation unit 15 adjusts the parameters of the VAV controller 1 by performing a trial operation of an air conditioning system including the air conditioner 3 and the VAV unit 4 (hereinafter referred to as “during adjustment”), the following formula (1) ) To obtain the airflow deviation normalized by the actual maximum airflow.
Airflow deviation = (Measured airflow-Required airflow) / Maximum airflow (1)

The air flow deviation calculation unit 15 standardizes the actual air flow using the following equation (2) when the air-conditioning system is used by the user after the air-conditioning system test operation (hereinafter referred to as “during operation”). Calculate the airflow deviation.
Airflow deviation = (measured airflow-correction required airflow) / capacity maximum airflow (2)

  The air volume control unit 16 generates a control signal indicating a damper opening for adjusting the air volume of the supply air from the VAV unit 4 in accordance with the air volume deviation calculated by the air volume deviation calculating unit 15. The air volume control unit 16 determines the damper opening by combining the two-position control on the cooling side and the two-position control on the heating side automatically by the well-known variable speed floating control, for example, Generate a signal.

  For example, as shown in FIG. 4, the air volume control unit 16 performs floating control of open output or closed output based on the positive / negative sign of the value of the air volume deviation (%) calculated by the air volume deviation calculation unit 15. The air volume control unit 16 performs variable control based on the air volume control step width corresponding to the magnitude of the air volume deviation value.

  The duct loss determination unit 17 determines the cause of the variation in the air volume of the supply air from the VAV unit 4. More specifically, the duct loss determining unit 17 is based on the required air volume calculated by the required air volume calculating unit 14 and the measured air volume acquired by the measured air volume acquiring unit 12 during the adjustment of the air conditioning system described above. The presence or absence of a defect in the duct 5 is determined.

  For example, when the required air volume is equal to the measured air volume in a predetermined time (minute) after the change of the required air volume, the duct loss determining unit 17 determines that there is no duct defect, and otherwise, the duct air loss determining unit 17 It is determined that there is a defect. When the duct defect determining unit 17 determines that there is a duct defect, a signal indicating the determination result is sent to the NN unit 18.

  Here, the duct defect indicates a duct pressure loss due to poor duct design or construction, and causes a variation in the amount of air supplied from the VAV unit 4. For this reason, if there is an individual difference in equipment such as a duct defect even if the required air volume that can reach the set temperature of the indoor area 6 in the shortest time is output in the feedback loop of the PID control by the required air volume calculation unit 14, As described above, the set temperature of the indoor area 6 may not be reached in the shortest time.

  The NN unit 18 learns the relationship between the required air volume and the measured air volume using a neural network, and estimates the correction amount of the required air volume using the learned neural network. More specifically, when it is determined that there is a duct defect at the time of adjustment of the air conditioning system, the NN unit 18 performs a neural operation on the difference between the calculated required air volume and the actual measured air volume affected by the duct defect. Learning using the network.

  The NN unit 18 extracts a feature quantity of the relationship between the required air volume and the measured air volume in order to estimate the correction amount of the required air volume so that the required air volume becomes the actual measured air volume (required air volume = measured air volume) Build a neural network. The feature value extracted by learning by the NN unit 18, that is, the parameter value such as the weight of the neural network is stored in the storage unit 20.

  Further, when the air conditioning system is operated, the NN unit 18 uses the required air volume calculated by the required air volume calculator 14 as an input, and estimates the required air volume correction amount using the learned neural network. More specifically, the NN unit 18 reads a parameter value such as a weight stored in the storage unit 20 and constructs a neural network. The NN unit 18 performs an operation on the learned neural network, outputs a correction amount of the required air volume, and obtains a corrected required air volume by correcting the required air volume based on the correction amount.

  As the neural network model, the NN unit 18 can use, for example, a well-known neural network model such as a recurrent neural network, a multilayer neural network, or a hierarchical Bayesian neural network.

  The output unit 19 outputs a control signal indicating the damper opening degree by the air volume control unit 16 to a damper opening adjusting mechanism (not shown) of the VAV unit 4. Based on this control signal, the opening degree of the damper 4-1 of the VAV unit 4 is adjusted.

  The storage unit 20 stores sensor data and calculation results necessary for control of the VAV controller 1. Specifically, the storage unit 20 stores updated values of parameters such as the set temperature of the indoor area 6, the measured indoor temperature, the required air volume, the measured air volume, the maximum air capacity, and the weight extracted by learning of the NN unit 18. Remember. Further, the storage unit 20 stores the correction amount, the correction required air volume, the value of the air volume deviation, the relationship between the air volume deviation used by the air volume control unit 16 and the damper output, and other setting information.

[Hardware configuration of VAV controller]
Next, the hardware configuration of the VAV controller 1 having the above-described functions will be described with reference to FIG.

  The VAV controller 1 includes a computer having an arithmetic device 102 having a CPU 103 and a main storage device 104 connected via a bus 101, a communication control device 105, an I / F 106, an external storage device 107, a display device 108, etc. It can be realized by a program that controls the hardware resources.

  The CPU 103 and the main storage device 104 constitute an arithmetic device 102. A program for the CPU 103 to perform various controls and calculations is stored in the main storage device 104 in advance.

  The communication control device 105 is a control device for connecting the VAV controller 1 and various external electronic devices via a communication network NW. A damper opening control signal may be sent to the VAV unit 4 via the communication control device 105 and the communication network NW. In addition, commands and control information from the air conditioning controller 2 may be received via the communication control device 105, or information related to control of the VAV unit 4 may be transmitted. Further, the set temperature of the indoor area 6 can be received via the communication control device 105.

  The I / F 106 is an interface for connecting the VAV controller 1 and various devices. An air speed sensor 8 for detecting the amount of air supplied from the VAV unit 4 and a temperature sensor 7 for detecting the indoor temperature of the indoor area 6 are connected via the I / F 106.

  The external storage device 107 includes a readable / writable storage medium and a drive device for reading / writing various information such as programs and data from / to the storage medium. For the external storage device 107, a semiconductor memory such as a hard disk or a flash memory can be used as a storage medium. The external storage device 107 is a sensor data storage unit 107a, a setting information storage unit 107b, an NN parameter storage unit 107c, a program storage unit 107d, and other storage devices (not shown), and is stored in the external storage device 107, for example. A storage device for backing up programs, data, and the like can be provided.

The sensor data storage unit 107a stores the indoor temperature, measured air volume, and the like of the indoor area 6 acquired from the sensors such as the temperature sensor 7 and the wind speed sensor 8 described with reference to FIG.
The setting information storage unit 107b stores setting values used for PID control by the required air volume calculation unit 14 described in FIG. 2 and variable speed floating control by the air volume control unit 16. In addition, the setting information storage unit 107b stores an initial value of the set temperature of the indoor area 6, initial setting information of a neural network model constituting the NN unit 18, and the like.

The NN parameter storage unit 107c stores parameter values such as weights of the neural network obtained as a learning result by the NN unit 18 described in FIG.
The sensor data storage unit 107a, the setting information storage unit 107b, and the NN parameter storage unit 107c correspond to the storage unit 20 described in FIG.

  The program storage unit 107d is required for the VAV controller 1 to control the air conditioning of the indoor area 6, such as the required air volume calculation process, the air volume deviation calculation process, the duct loss determination process, the learning process, and the estimation process in the present embodiment. Various programs for executing arithmetic processing are stored.

  The display device 108 is realized by a liquid crystal display or the like. The display device 108 may display sensor information such as the measured air volume acquired by the VAV controller 1 and the opening degree of the damper 4-1.

[Operation of VAV controller]
Next, the operation of the VAV controller 1 having the above-described configuration will be described with reference to the flowcharts of FIGS. In the present embodiment, the air conditioning system is trial run in a new building to adjust the parameters of the VAV controller 1 and the like, and after the trial run, the air conditioning system is operated for use by the user. The operation of the VAV controller 1 will be described separately.

[Operation of VAV controller during adjustment]
FIG. 6 is a flowchart for explaining the overall operation of the VAV controller 1 during adjustment.
First, the required air volume calculation unit 14 performs PID control based on the set temperature and the room temperature, and calculates the required air volume (step S1). Next, the NN unit 18 learns the relationship between the required air volume calculated in step S1 and the actual measured air volume using a neural network (step S2). At this time, the air volume control unit 16 performs variable speed floating control according to the air volume deviation calculated based on the required air volume calculated in step S1, the actual measured air volume, and the actual maximum air volume, and the damper. A control signal indicating the opening is generated (step S3).

[Required air volume calculation processing]
FIG. 7 is a flowchart for explaining the required air volume calculation process (step S1).
First, the set temperature acquisition unit 10 acquires a set temperature in the indoor area 6 to be controlled (step S10). Next, the room temperature acquisition unit 11 acquires the room temperature from the temperature sensor 7 (step S11). The acquired set temperature and indoor temperature of the indoor area 6 are stored in the storage unit 20.

  Next, the actual maximum air volume acquisition unit 13 acquires the measurement air volume when the air conditioner 3 has the maximum air volume and the VAV unit 4 is fully opened from the measurement air volume acquisition unit 12 (step S12). Is stored in the storage unit as the maximum actual air volume.

  Next, the required air volume calculation unit 14 performs PID control using the set temperature of the indoor area 6 acquired in step S10 as a target value, the indoor temperature acquired in step S11, and the required air volume as an operation amount (step S13). ). More specifically, the required air volume calculation unit 14 reads out the minimum air volume set in advance in the shipping inspection or the like by the manufacturer of the VAV unit 4 from the storage unit 20, and requests based on the actual maximum air volume acquired in step S12. Calculate the air volume.

  The required air volume calculated by the required air volume calculating unit 14 is stored in the storage unit 20 (step S14).

  Next, a learning process (step S2) and an air volume control process (step S3) executed in parallel after the required air volume calculation process (step S1) will be described with reference to the flowcharts of FIGS.

[Learning process]
As illustrated in FIG. 8, the NN unit 18 reads out the required air volume calculated by the required air volume calculating unit 14 from the storage unit 20 (step S20). Next, the measured air volume acquisition unit 12 acquires the measured air volume converted based on the wind speed detected by the wind speed sensor 8 (step S21).

  Next, the duct loss determination unit 17 determines the presence or absence of a defect in the duct based on the required air volume and the measured air volume. If there is a duct defect (step S22: YES), the neural network calculation by the NN unit 18 is performed. Is executed (step S23).

  More specifically, the duct defect determining unit 17 determines that a duct defect exists when the required air volume does not become the measured air volume in a predetermined time (minutes) after the change of the required air volume.

  Further, the NN unit 18 learns the relationship between the required air volume and the measured air volume using a preset neural network model, and calculates a feature quantity for estimating the required air volume correction amount considering the duct defect. Extract. More specifically, the NN unit 18 uses the required air volume calculated by the required air volume calculator 14 and the actual measured air volume as teacher data. The NN unit 18 adjusts the parameters of the neural network so as to output a correction amount of the required air volume so that the actual measured air volume is the same as the required air volume (requested air volume = measured air volume) in response to the input of the required air volume. To do.

  The NN unit 18 repeatedly updates parameter values such as weights and biases so that the error between the output value (correction amount) of the neural network with respect to the input data of the required airflow for learning and the teacher data decreases to an allowable range. . The NN unit 18 stores the finally determined parameter update value in the storage unit 20 (step S24). A learned neural network that estimates the correction amount of the required air volume (corrected required air volume) based on the parameter value determined by learning in this way is constructed.

  The neural network model constituting the NN unit 18 is not limited to using supervised learning, and unsupervised learning or semi-supervised learning may be used.

[Air volume control during adjustment]
Next, the air volume control process executed in parallel with the learning process at the time of adjustment will be described with reference to the flowchart of FIG.
First, the air volume control unit 16 reads the required air volume stored in the storage unit 20 (step S30). Next, the measured air volume acquisition unit 12 acquires the measured air volume converted based on the wind speed detected by the wind speed sensor 8 (step S31).

  Next, the air volume control unit 16 reads the actual maximum air volume stored in the storage unit 20 (step S32). Thereafter, the air volume deviation calculation unit 15 calculates the air volume deviation using the above-described equation (1) based on the required air volume, the measured air volume, and the actual maximum air volume (step S33). The calculated airflow deviation is temporarily stored in the storage unit 20.

  Thereafter, the air volume control unit 16 obtains a damper opening degree of the VAV unit 4 based on the calculated air volume deviation by variable speed floating control, and generates a control signal indicating the damper opening degree (step S34). And the air volume control part 16 sends out the control signal which shows the produced | generated damper opening degree to the damper opening degree adjustment mechanism of the VAV unit 4 via the output part 19 (step S35).

  Thus, at the time of adjustment of the air conditioning system, the required air volume calculated by the required air volume calculating unit 14 is used for both the air volume control process and the learning process, and these processes are executed in parallel.

[Operation of VAV controller during operation]
Next, the operation of the VAV controller 1 during operation of the air conditioning system will be described using the flowcharts of FIGS. 10 to 12.
FIG. 10 is a flowchart for explaining the overall operation of the VAV controller 1 during operation. When each process at the time of adjustment of the air conditioning system described in FIGS. 6 to 9 is completed, the air conditioning system is operated by the user this time.

  As shown in FIG. 10, the required air volume calculation unit 14 calculates the required air volume in the same manner as during adjustment (step S1). Next, the NN unit 18 estimates a correction amount for the calculated required air volume using the learned neural network (step S2A). Thereafter, the air volume control unit 16 obtains the damper opening degree using the correction required air volume corrected with the correction amount estimated in step S2A (step S3A).

[Guess process]
Here, the estimation process (step S2A) executed by the NN unit 18 will be described with reference to the flowchart of FIG. The required air volume calculation process (step S1) is the same as the process at the time of adjustment, and the description thereof is omitted.

  First, the NN unit 18 reads the required air volume stored in the storage unit 20 (step S200). Next, the NN unit 18 reads the parameter value of the learned neural network stored in the storage unit 20 (step S201). The NN unit 18 loads parameter values to construct a neural network.

  Thereafter, the NN unit 18 uses the learned neural network to perform a neural network calculation using the required air volume as an input (step S202). The NN unit 18 estimates a required air volume correction amount in consideration of variations in the air volume of the VAV unit 4 due to a duct defect by a neural network calculation. The NN unit 18 calculates the correction required air volume based on the correction amount. The calculated correction required air volume is stored in the storage unit 20 (step S203).

[Airflow control processing during operation]
Next, the air volume control process (step S3A) executed after the estimation process during the operation of the air conditioning system will be described with reference to the flowchart of FIG.
First, the air volume control unit 16 reads the correction required air volume obtained by the NN unit 18 from the storage unit 20 (step S300). Next, the measured air volume acquisition unit 12 acquires the measured air volume converted based on the wind speed detected by the wind speed sensor 8 (step S301).

  Next, the air volume control unit 16 reads the actual maximum air volume stored in the storage unit 20 (step S302). After that, the air volume deviation calculation unit 15 calculates the air volume deviation using the above-described equation (2) based on the measured air volume, the actual maximum air volume, and the correction required air volume (step S303).

  Thereafter, the air volume control unit 16 obtains the damper opening of the VAV unit 4 by variable speed floating control based on the air volume deviation calculated in step S303, and generates a control signal (step S304). The air volume control unit 16 sends the generated control signal to the damper opening adjustment mechanism of the VAV unit 4 via the output unit 19 (step S305). Based on the control signal indicating the damper opening, the opening of the damper 4-1 of the VAV unit 4 is adjusted, and the air volume of the supply air is adjusted, so that the room temperature in the indoor area 6 approaches the set temperature. become.

  As described above, according to the VAV controller 1 according to the present embodiment, at the time of adjusting the air conditioning system, the relationship between the calculated required air volume and the actual measured air volume affected by the duct defect is learned. . Further, during the operation of the air conditioning system, a correction required air volume (correction amount) that complements the variation in the air volume due to the duct defect is estimated, and a control signal indicating the duct opening degree is generated based on the correction required air volume. Therefore, the VAV controller 1 can cause the indoor temperature of the indoor area 6 to reach the set temperature in a shorter time.

  In addition, the VAV controller 1 according to the present embodiment includes an NN unit 18 that performs a neural network operation, and learning processing is automatically performed inside the VAV controller 1. For this reason, when the air conditioning system is adjusted, there is no need for new work by humans, and more efficient operation of the air conditioning system is realized.

  The embodiments of the air conditioning control device and the air conditioning control method of the present invention have been described above. However, the present invention is not limited to the described embodiments, and those skilled in the art can assume the scope of the invention described in the claims. Various possible modifications can be made.

  For example, in the embodiment described above, the case where the VAV controller 1 is provided for each indoor area 6 to be controlled and the NN unit 18 performs learning and estimation for each indoor area 6 has been described. However, for example, a predetermined one of the plurality of VAV controllers 1 collects information on learning results of the NN unit 18 in each of the plurality of indoor areas 6 to be controlled, and corrects the fan rotation speed of the air conditioner 3. May be. By adopting such a configuration, more comfortable air conditioning control can be realized.

  In the embodiment described above, the case where the VAV controller 1 corrects the required air volume in consideration of the duct defect has been described. However, it is not limited to a duct defect as long as it is an individual difference of devices that affects the arrival time to the target value in the feedback loop of PID control.

  DESCRIPTION OF SYMBOLS 1 ... VAV controller, 2 ... Air conditioning controller, 3 ... Air conditioner, 4 ... VAV unit, 4-1 ... Damper, 5 ... Duct, 6 ... Indoor area, 7 ... Temperature sensor, 8 ... Wind speed sensor, 9 ... Outlet, DESCRIPTION OF SYMBOLS 10 ... Setting temperature acquisition part, 11 ... Indoor temperature acquisition part, 12 ... Measurement air volume acquisition part, 13 ... Ability maximum air volume acquisition part, 14 ... Required air volume calculation part, 15 ... Air volume deviation calculation part, 16 ... Air volume control part, 17 DESCRIPTION OF SYMBOLS Duct defect | deletion determination part, 18 ... Neural network (NN) part, 19 ... Output part, 20 ... Memory | storage part, 101 ... Bus, 102 ... Arithmetic apparatus, 103 ... CPU, 104 ... Main memory, 105 ... Communication control apparatus, 106 ... I / F, 107 ... external storage device, 107a ... sensor data storage unit, 107b ... setting information storage unit, 107c ... NN parameter storage unit, 107d ... program storage unit, 108 Display device.

Claims (8)

An air conditioning control device for controlling a variable air volume unit provided for each controlled area,
A required air volume calculation unit configured to calculate a required air volume, which is an air volume required for supplying air from the variable air volume unit, so that the indoor temperature of the controlled area reaches a set temperature;
A control signal indicating a duct opening for adjusting the air volume of the supply air from the variable air volume unit is generated according to a deviation between the required air volume and the measured air volume of the air supply from the variable air volume unit. An air flow control unit,
A neural network unit that learns the relationship between the required air volume and the measured air volume using a neural network, and constructs a learned neural network that estimates a correction amount of the required air volume;
With
The neural network unit estimates the correction amount using the learned neural network,
The air volume control unit adjusts the air volume of the supply air from the variable air volume unit according to the deviation between the corrected required air volume corrected according to the estimated correction amount and the measured air volume. An air conditioning control device characterized by generating a control signal indicating In the air-conditioning control device according to claim 1,
A determination unit configured to determine whether or not there is a factor that causes variation in the air volume of the air supply from the variable air volume unit;
The neural network unit learns a relationship between the required air volume and the measured air volume using a neural network when the determining unit determines that the factor is present. In the air conditioning control device according to claim 1 or 2,
The air volume control unit performs variable speed floating control according to the deviation and generates the control signal. In the air-conditioning control device according to any one of claims 1 to 3,
The air-conditioning control apparatus further comprising a first deviation calculating unit configured to calculate the deviation between the required air volume and the measured air volume. In the air-conditioning control device according to any one of claims 1 to 4,
The air-conditioning control apparatus further comprising a second deviation calculating unit configured to calculate the deviation between the correction required air volume and the measured air volume. In the air-conditioning control device according to any one of claims 1 to 5,
The required air volume calculating unit calculates the required air volume by PID control using the set temperature as a target value, the room temperature as an input value, and the required air volume as an operation amount. An air conditioning control method for controlling a variable air volume unit provided for each controlled area,
A required air volume calculating step for calculating a required air volume required for the air volume of the supply air from the variable air volume unit so that the indoor temperature of the controlled area reaches a set temperature;
An air volume control step of generating a control signal indicating a duct opening for adjusting the air volume of the air supply from the variable air volume unit according to a deviation between the required air volume and the measured air volume of the air supply from the variable air volume unit When,
A determination step of determining the presence or absence of a factor that causes variation in the air volume of the supply air from the variable air volume unit;
A learning step of learning, using a neural network, a relationship between the required air volume and the measured air volume of the supply air from the variable air volume unit when it is determined in the determining step that the factor is present;
A storage step of storing in the storage unit information relating to the learned neural network constructed in the learning step. An air conditioning control method for controlling a variable air volume unit provided for each controlled area,
A required air volume calculating step for calculating a required air volume required for the air volume of the supply air from the variable air volume unit so that the indoor temperature of the controlled area reaches a set temperature;
A guessing step of guessing a correction amount of the required air volume using a learned neural network stored in a storage unit;
The air volume of the supply air from the variable air volume unit is adjusted according to the deviation between the correction required air volume corrected according to the correction amount estimated in the estimating step and the measured air volume of the air supply from the variable air volume unit. An air volume control step for generating a control signal indicating a duct opening degree for performing the air conditioning control method.