JP2001221481A - Controller for air conditioning - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a controller for air conditioner in which the temperature of an objective space can be regulated quickly and easily. SOLUTION: A temperature regulator 32 employing a previously learnt neural network having the temperature in a large space detected by a temperature sensor 16, a set temperature of the large space set by an input unit 30, and a heat generation rate in the large space measured by a radiation temperature sensor 28 as inputs and the control value of an air conditioner as an output operates a set temperature in the vicinity of the output of the air conditioner 12 such that the future temperature in the large space has a set value. Based on the operated outlet temperature of the air conditioner, output of the air conditioner is regulated and the indoor temperature is controlled.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an air-conditioning control device, and in particular, when a fluid is supplied from an air conditioner to a target space, a fluid whose temperature has been adjusted based on at least the temperature of the target space is output from the air conditioner. For this purpose, the present invention relates to an air conditioning control device for controlling the air conditioning equipment.

[0002]

2. Description of the Related Art In a living space, the temperature is often adjusted by air conditioning in order to make the occupants comfortable. In general, the temperature condition in the living space is monitored by a temperature sensor, and the output of the air conditioner (the valve of the cold / hot water) is maintained so that the temperature in the living space is kept constant (to eliminate the difference from the set temperature). Opening, etc.).

In this case, feedback control is performed so as to eliminate the difference between the temperature indicated by the temperature sensor in the living space and the set temperature in the living space. At this time, the temperature in the living space changes due to fluctuations in the outside air temperature, disturbances such as the presence or absence of insolation, the number of occupants, increase / decrease in heat generation of equipment installed in the living space, and the amount of heat supplied from the air conditioner. . Therefore, in the feedback control, the control is performed based on the temperature in the living space that appears as a result thereof.

In a relatively small living space, it is possible to maintain a constant temperature in the living space by air conditioning control by feedback control. However, in a large space such as a hall or an event hall, a large amount of heat is supplied. Is generated by the air conditioner,
It needs to be supplied in a large space. For this reason, as shown in FIG. 1, an air conditioner 12 is provided outside the large space 10, and a fluid such as air whose temperature is adjusted from the air conditioner 12 to the inside of the large space 10 by a heat transfer path (air conditioning air supply duct) 14. (Less than,
SA (referred to as an air-conditioning fluid) is conveyed and supplied to a large space. Then, in order to maintain the temperature in the large space 10 at the set value, the temperature in the large space 10 is detected by the temperature sensor 16 and the control logic 18 supplies the air-conditioning fluid to the air conditioner 12 according to the detected temperature. Valve 20 to control
The degree of opening is adjusted.

However, due to the length of the air-conditioning air supply duct, the distance from the air-conditioning outlet to the living area, etc., there is a time lag when the result of adjusting the output of the air conditioner reaches the temperature in the living space. It cannot be ignored in large-scale air conditioning such as.

Therefore, as shown in FIG. 2, a cascade control of a large space temperature → (primary temperature controller) → air conditioner outlet temperature → (secondary temperature controller) → air conditioner output is generally used. ing. That is, the primary temperature controller 24 and the secondary temperature controller 26 are connected in cascade as control logic (a so-called cascade PID control method in which feedback is performed from the internal temperature via the temperature immediately after the air conditioner).
In the primary temperature controller 24, the temperature in the large space 10 is detected by the temperature sensor 16, and a target value of the temperature immediately after the air conditioner with respect to the detected temperature is output. The secondary temperature controller 26 detects the temperature immediately after the output of the air conditioner by the immediately after temperature sensor 22 and adjusts the opening degree of the valve 20 so as to reach the target value of the input temperature immediately after the air conditioner. . By doing so, the responsiveness of the air conditioning is improved.

[0007]

However, disturbances such as outside air temperature and solar radiation, and increase and decrease in heat generation of personnel and equipment gradually appear in temperature changes in the living space. Therefore, if the air conditioner is adjusted only by the temperature in the living space, the temperature fluctuation increases due to disturbance, the temperature fluctuation increases due to heat generation fluctuation, and the temperature fluctuation width increases. In particular, in a large space, the temperature change appears more slowly. Therefore, if only the temperature in the living space is monitored, the output of the air conditioner cannot be adjusted in time, and the temperature in the living space, that is, in the large space, fluctuates.

For example, in a large space, if the number of personnel suddenly decreases during cooling, the output of the air conditioner changes only gradually.
The air conditioner was supplied with too much cold heat, resulting in excessive cooling in the living space.

An object of the present invention is to provide an air-conditioning control device capable of quickly and easily adjusting the temperature of a target space in consideration of the above fact.

[0010]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a method for controlling an air conditioner to output a fluid whose temperature has been adjusted from the air conditioner when the fluid is supplied from the air conditioner to a target space. Detecting means for detecting a temperature in a target space, measuring means for measuring a heat source in the target space, setting means for setting a temperature in the target space, and a temperature detection value in the target space. Input means for inputting a heat source measurement value in the target space and a temperature set value in the target space; output means for outputting a control value of the air conditioner; and transmitting data from the input means to the output means. Between the input means and the output means with a neural network model having a multilayer structure in which connection relations of a large number of neural circuit elements are modeled. A temperature detection value, a heat source measurement value in the target space and a control value of the air conditioner corresponding to a temperature set value in the target space, between the input means and the intermediate means, and the intermediate means Correspondence means for determining a connection relationship of the neural circuit element with the output means in advance.

First, the temperature in the target space is detected by the detecting means, and the heat source in the target space is measured by the measuring means.
The temperature in the target space is set by setting means. Then, the corresponding unit outputs a control value of the air conditioner corresponding to the detected temperature value, the measured heat source value, and the set temperature value in the target space. The corresponding unit is an input unit for inputting a temperature detection value, a heat source measurement value, and a temperature set value in the target space, an output unit for outputting a control value of the air conditioner, and transmitting data from the input unit to the output unit. Intermediate means for Between these input means and the output means, a neural network model (a so-called neural network) having a multilayer structure in which a connection relation of a large number of neural circuit elements is modeled. In this neural circuit model, between the input means and the intermediate means, and between the intermediate means and the output means, so as to output the detected value of the temperature in the target space, the control value of the air conditioner corresponding to the heat source measurement value and the temperature set value. Are predetermined in advance. Thus, the air conditioner can be quickly controlled according to the temperature fluctuation in the target space, and the temperature fluctuation in the target space can be suppressed.

The response means includes a plurality of the detected temperature values,
The connection between the neural circuit elements included in the input means to the output means of the corresponding means, using the relation data in which the correspondence between the heat source measurement value and the temperature set value and a number of control values of the air conditioner is known in advance. Can build relationships. By using a large number of known relationship data, more practical temperature detection value in the target space, heat source measurement value and temperature set value,
Correspondence with the control value of the air conditioner can be established.

[0013] The measuring means may measure the total number of persons and the total amount of heat existing in the target space. That is, although the target space includes heat sources such as occupants and machines, there is a large number of variances in the object space. In this case, if the total number of persons existing in the target space is measured, the heat source in the target space can be easily estimated, and the amount of heat generated by the heat source can be estimated. Further, the total number of heat sources such as machines may fluctuate, or the amount of heat of the heat sources may fluctuate. In this case, the measuring means may measure the total number of heat sources such as machines and the amount of heat of the heat sources.

[0014] The measuring means may measure the distribution of persons existing in the target space. If the distribution of persons existing in the target space is measured, when there is a local temperature change, the local temperature change can be easily reflected in the target space.

The air-conditioning control device further comprises a detecting means for detecting an output temperature of the air-conditioning equipment, and the corresponding means causes the input means to further input a detected value of the output temperature of the air-conditioning equipment, and causes the output means to output the target temperature. The time to reach the temperature set point in the space can be predicted.

A detecting means for detecting the output temperature of the air conditioner is further provided. In the corresponding means, a detected value of the output temperature of the air conditioner is further inputted, and the set temperature in the target space is reached from the output means. If the time is output, the time to reach the temperature set value in the target space can be predicted.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the present embodiment, the present invention is applied to an air-conditioning control device that adjusts and controls the temperature in a large space of a dome-shaped space as a target space to a set temperature.

In a large space with a high personnel density, such as a dome or a hall, the heat load from personnel is a significant factor in disturbance. Therefore, in order to mitigate the effect of the fluctuation and improve the responsiveness, a method of applying a neural network to the primary temperature controller of the cascade control and performing predictive control (hereinafter, NN applied cascade control) is effective. Was obtained. In the present embodiment, the internal heat generation is intended to be reflected in the calculation of the control output (set temperature immediately after the air conditioner) in a feedforward manner in addition to the input items of the primary temperature controller.

As shown in FIG. 3, in the present embodiment, an air conditioner 12 is provided outside the large space 10. The air conditioner 12 includes a blower 12A and a hot / cold coil 12B. The air conditioner 12 and the large space 10 are communicated as an air supply side by an air conditioning air supply duct 14 and are communicated as an exhaust side by an air conditioning exhaust duct 15. By operating the blower 12A, the air-conditioning fluid in the air conditioner 12 is supplied to the large space 10 through the air-conditioning air supply duct 14, and the large space 1
The air inside 0 is returned to the air conditioner 12 through the air conditioning exhaust duct 15.

When the air in the air conditioner 12 is output to the air conditioning air supply duct 14 by the operation of the blower 12A, the heating / cooling coil 1
2B. This heating / cooling coil 12
B is connected to the transport path 13. A catalyst (refrigerant or hot medium) having a predetermined temperature is circulated in the transfer path 13. By adjusting the opening of a valve 20 provided in the middle of the transfer path 13, The configuration is such that the temperature of the air passing through the hot / cold coil 12B is adjusted and supplied to the air conditioning air supply duct 14 as an air conditioning fluid.

A secondary temperature controller 26 is connected to the valve 20. The secondary temperature controller 26 is connected to the temperature sensor 22 provided immediately after the air-conditioning air supply duct 14 on the side of the hot / cold coil 12B and the NN applied primary temperature controller 32 to which the present invention is applied. I have. The NN applied primary temperature controller 32 has an input device 30 for inputting a set temperature,
A measuring device 28 for measuring heat generation and a temperature sensor 16 for detecting a temperature in the large space 10 are connected.

As shown in FIG. 4, the large space 10 is a dome-shaped living space, and has an upper intake port * 1 and an upper floor discharge slit as portions communicating with the air conditioning air supply duct 14 and the air conditioning exhaust duct 15. * 2, lower outlet * 3, lower inlet standing seat * 4, lower inlet concourse * 5, lower inlet field * 6.

Measuring device 2 for measuring heat generation in large space 10
8 has a radiation temperature sensor. What is necessary is just to measure the heat generation in a large space by this radiation sensor. In addition, the heat generation can be obtained by calculation without directly measuring the heat generation. For example, a standard calorific value J of the human body P is prepared in advance as data, the number of people N resident in the large space 10 is measured, the product of the calorific value J and the number of people N is obtained, and the calculation result is calculated as the total calorific value. Z (= J × N) may be set. In this case, large space 1
In order to accurately measure the number of people N living in 0, it is preferable to measure the total number of people at the current time by using a passer counter device that counts entrances and exits, a ticket gate for entrance tickets, and the like.

The measuring device 28 is not limited to measuring only the total number of human bodies P or the total amount of heat. For example, any device can be used as long as it can measure a heat source relating to a person who is predicted to generate heat fluctuation in a large space. For example, the total number of devices and the amount of heat that generate heat may be measured, and the total number and the amount of heat may be measured for each of a person, that is, a person and a device.

NN applied primary temperature controller 3 of this embodiment
Numeral 2 detects the temperature and heat generation in the large space, and outputs a control value for performing air-conditioning control in a feed-forward manner on the influence of these on the future. More specifically, as shown in FIG. 5, the NN applied primary temperature controller 32 has a conversion function configured by a conceptual neural network and also has a learning function of learning the conversion function. In addition,
The correspondence between the temperature and heat value in the large space and the control value for air conditioning control is learned in advance by another neural network,
By inputting the learned conversion coefficients of another neural network, a neural network may be constructed using the conversion coefficients.

More specifically, the NN applied primary temperature controller 32
Has an input layer having neurons corresponding to the number of temperatures and heat values in the large space as input layers to enable input of each value of temperature and heat value in the large space, and air conditioning as an output layer via an intermediate layer. The neural network has one neuron (one in this embodiment) corresponding to the number of control values for control, and forms a neural network in which each neuron is connected by a synapse. When each value of the large space temperature and the calorific value is input to the NN applied primary temperature controller 32 after learning, which will be described later, a control value for controlling air conditioning corresponding to the input value is output. At the time of learning, the temperature immediately after the air conditioner 12 is used as a control value corresponding to the temperature in the large space and the calorific value (see FIG. 7).
Is input as a teacher, and the value of the temperature and the amount of heat generated in the large space are set to correspond to the control value according to the magnitude of an error difference between the control value of the output and the known control value.

FIG. 5 shows an example of the neural network used in the NN applied primary temperature controller 32.
, An input layer composed of a predetermined number of units I1, I2,..., Ip (p> 1) corresponding to neurons,
.., Mq (q> 1), and an output layer including a predetermined number (one in the present embodiment) of output units U1.
In the present embodiment, the output units are connected to the offset units 46 and 48 for offsetting the output value by a predetermined value to each unit of the intermediate layer and the unit of the output layer. Is not provided.

The unit of the intermediate layer and the unit of the output layer are constituted by neural circuit elements having a sigmoid characteristic whose input / output relationship is represented by a sigmoid function, and the unit of the input layer is constituted by a neural circuit element having a linear input / output relationship. Have been. By configuring so as to have this sigmoid characteristic, the output value becomes a real value (positive number).

Each output of the unit of the intermediate layer is a virtual internal unit corresponding to the average value of the membrane potential of the neuron, using a weight (coupling coefficient of the unit) corresponding to the strength of each synaptic connection on the input side as an input signal. It can be represented by a non-linear function that represents the characteristics of the neuron by the state variable. The same applies to each output of the unit in the output layer. Note that the characteristics of each unit in the input layer may be characteristics that output the input as it is. The weights (coupling coefficients) of these units are learned and corrected by learning processing to be described later so that errors in known learning data are minimized. That is, it is thought that the nervous system of an organism learns by changing the strength of synaptic connections between neurons. For this reason,
In the neural network, learning / correction refers to changing the weight (coupling coefficient) of each unit corresponding to the synapse weight (see FIG. 6 for the learning result).

Next, the details of the neural network learning process in the NN applied primary temperature controller 32 will be described with reference to FIG. Here, the air conditioner 12 is actually operated based on the set temperature, the large space temperature, and the heat value, and a number of control values of the secondary temperature controller 26 at that time are obtained in advance. The correspondence between the set temperature, the large space temperature, and the calorific value and the control value of the secondary temperature controller 26 is used as learning data, and a plurality of teacher data used at the time of learning. During this learning, the temperature immediately after the air conditioner 12 may be measured by the temperature sensor 22 immediately afterward, and the value may be used as a teacher (see FIG. 7).

First, in step 200,
Read the training data. In the next step 202, initialization is performed by setting the coupling coefficient (weight) of each unit in the neural network to a predetermined value. In the next step 204, an error of each unit of the intermediate layer and the output layer is obtained in order to train the neural network using the plurality of known learning data. Here, the error of the output layer, that is, the error of the unit can be minimized by slightly changing each coupling coefficient. The error of the middle layer is
It can be obtained by an inverse calculation such as an error back propagation method (a so-called back propagation method) using the error of the output layer.

In the next step 206, the obtained coupling coefficients are updated (rewritten), and in the next step 208, it is determined whether or not the error of the control value is within a predetermined range which is a criterion for convergence determination. Thus, it is determined whether or not convergence has been achieved, or whether or not the above processing has been repeated a predetermined number of times. If the determination is affirmative, this routine ends.
On the other hand, if the determination is negative, the process returns to step 204, and the above processing is repeated. In this way, when learning data is input, each coupling coefficient and threshold value are determined so that the error of each unit of the intermediate layer and the output layer is minimized.

In this way, the neural network is trained using a plurality of known learning data.

In this way, the temperature in the large space 10, the set temperature, and the amount of heat generated in the large space 10 are input, and a temperature controller using a neural network is applied so that the future indoor temperature becomes a set value. Calculate the set temperature of the air conditioner outlet. The output of the air conditioner is adjusted based on the calculated air conditioner outlet temperature to control the indoor temperature.

It is to be noted that, without providing the secondary temperature controller 26, the room temperature, the air conditioner outlet temperature, and the indoor heat generation are input and the temperature controller 32 using a neural network is input.
Only with this, the output of the air conditioner may be directly adjusted so that the future indoor temperature becomes the set value.

In the above embodiment, the air conditioner is controlled based on the temperature in the large space and the amount of heat generated in the large space. However, the present invention is not limited to this, and furthermore, predictive control is possible. . That is, each value obtained by adding the temperature in the large space, the calorific value in the large space, and the air conditioner outlet temperature is input and N
By using the temperature reached in the large space 10 in the future (that is, after a certain period of time or in a steady state) as the output of the neural network of the N applied primary temperature controller 32, the current temperature in the large space, From the calorific value and the air conditioner outlet temperature, it is possible to predict and calculate the future temperature in real time. Then, the NN applied primary temperature controller 32 further calculates the set temperature of the air conditioner outlet such that the temperature inside the large space 10 in the future will become the set value. The output of the air conditioner 12 is adjusted on the basis of the predicted future room temperature to control the large space temperature. By doing so, the large space 10
It is possible to shorten the time required for the temperature inside to reach the set temperature, and to suppress the fluctuation range of the temperature.

For example, as shown in FIG.
The secondary temperature controller 32 includes an NN primary controller 32A and a controller 32B. On the input side of the NN primary controller 32A, a temperature sensor 16 for detecting the temperature in the large space 10,
A measuring device 28 for measuring heat generation and a temperature sensor 22 for detecting a temperature immediately after the output of the air conditioner 12 are connected. The NN applied primary temperature controller 32 outputs the future temperature of the large space, and its output side is connected to the input side of the controller 32B. An input device 30 for inputting a set temperature of a large space is further connected to an input side of the controller 32B. The controller 32B outputs the set temperature of the air conditioner 12, and its output side is connected to the input side of the secondary temperature controller 26. An immediate temperature sensor 22 is further connected to the input side of the controller 32B.

In this manner, the future temperature of the large space is obtained inside the NN applied primary temperature controller 32, and the set temperature of the air conditioner is obtained using the temperature. If the air conditioner is controlled based on the set temperature and the outlet temperature of the air conditioner, the time until the temperature in the large space 10 reaches the set temperature can be shortened, and the fluctuation range of the temperature can be suppressed.

In FIG. 10, the NN applied primary temperature controller 32 is constituted by the NN primary controller 32A and the controller 32B. However, only the neural network in the NN applied primary temperature controller 32 is used. It is also possible to configure.
In this case, as shown in FIG. 11, the NN applied primary temperature controller 32 has a temperature sensor 1 for detecting the temperature in the large space 10.
6. Measuring device 28 for measuring heat generation, and air conditioner 1
The temperature immediately after the output of No. 2 is connected to the temperature sensor 22 immediately after the output.
The NN application primary temperature controller 32 outputs the future temperature reached in the large space, and its output side is connected to the input side of the secondary temperature controller 26. An input device 30 for inputting a set temperature of a large space is further connected to an input side of the secondary temperature controller 26. The secondary temperature controller 26 obtains an output value from the set temperature of the air conditioner 12 and the temperature reached in the future, and controls the air conditioner.

In this case, since the air conditioner is controlled based on the set temperature of the air conditioner obtained from the future temperature of the large space and the target set temperature, the temperature in the large space 10 reaches the set temperature. The time required to perform the process can be shortened, and the fluctuation range of the temperature can be suppressed.

As described above, in the present embodiment, in addition to the temperature in the large space, the heat generation is detected and added to the control input items, and the air-conditioning control (feed-forward-like) that anticipates their influence on the future is performed. ing. That is, for air conditioning control,
Using a neural network (similar to multiple regression analysis), which is an operation of multiple inputs and one output, the relationship between the temperature state in the large space 10 and the air-conditioning control value, which is a non-linear correspondence, is derived. Since the air-conditioning control is performed in advance of the influence on the future, the fluctuation of the room temperature can be suppressed even when the heat generation changes.

Therefore, in addition to the detected temperature in the large space, the amount of heat generated in the large space is input, and the air conditioning control taking into account the effect of the heat generated in the large space on the future is performed. Fluctuations in the internal temperature can be suppressed. Thereby, the accuracy of temperature management in a large space can be improved. Also, it can contribute to energy saving.

FIG. 9 shows the results of a comparative study of the response with the fixed supply air temperature, the response with the direct PID control method, and the response with the NN applied cascade control method to which the present invention is applied. In this comparative study, a 200 m diameter dome having a transparent part for natural daylight at the roof top was assumed as the large space shown in FIG. And the supply air temperature was adjusted.

First, when the supply air temperature is fixed, immediately after the internal load decreases (the number of personnel decreases), the temperature suddenly drops.
After that, the gradient gradually decreases and becomes constant. In addition, P
When the ID control is performed, immediately after the internal load decreases (the number of personnel decreases), the temperature decreases with a gentler gradient than the fixed supply air temperature, and gradually approaches the set temperature from the decrease. On the other hand, when the NN application cascade control to which the present invention is applied is performed, immediately after the internal load decreases (the number of personnel decreases), the PI
The temperature decreases at a gentler gradient than in the D method, and gradually approaches the set temperature at a time earlier than the direct PID method from the decrease. Furthermore, the present invention is applied to N
When the N applied cascade control is performed, the range of the temperature fluctuation is suppressed, so that the temperature can be adjusted with high accuracy.

As described above, according to the present embodiment, although there is some fluctuation immediately after the load depending on the load, the temperature can be controlled to the set temperature earlier than other methods. Therefore, in the present embodiment, it is shown that the control output and the supply air temperature are adjusted and the fluctuation of the internal temperature is suppressed.
Further, since the range of the temperature fluctuation can be suppressed, the temperature can be adjusted accurately.

In this embodiment, the case where the present invention is applied to a large space such as a hall as a target space has been described, but the target space of the present invention is not limited to a large space. For example, the present invention is effective in a space in which heat is generated from a device, a space which is affected by temperature due to disturbance such as solar radiation or an outside temperature, and a space in which heat is fluctuated. Further, since the range of the temperature fluctuation can be suppressed, the temperature can be adjusted accurately.

[0047]

As described above, according to the present invention, since the heat generation in the target space is taken into consideration in addition to the temperature in the target space, it is possible to control the air-conditioning taking into account their influence on the future. In addition, there is an effect that the fluctuation of the temperature in the target space can be suppressed even when the heat generation changes.

Further, by further considering the temperature at the outlet of the air conditioner, there is an effect that the set temperature attainment in the target space can be predicted.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a conceptual configuration of a general air conditioning control system.

FIG. 2 is a block diagram illustrating a conceptual configuration of a cascade PID control method.

FIG. 3 is a block diagram illustrating a schematic configuration of an air conditioning control device according to the present embodiment.

FIG. 4 is an explanatory diagram showing an example of a large space.

FIG. 5 is a block diagram showing a configuration of a conceptual neural network included in the NN applied primary temperature controller 32.

FIG. 6 is a block diagram showing a configuration of a conceptual neural network included in the NN applied primary temperature controller 32 after learning.

FIG. 7 is a block diagram illustrating a conceptual configuration of an air conditioning control device according to the present embodiment.

FIG. 8 is a processing routine showing a flow of processing of learning of the neural network.

FIG. 9 is a characteristic diagram showing the results of a comparative study of the response by the fixed supply air temperature, the response by the direct PID control method, and the response by the NN applied cascade control method.

FIG. 10 is a block diagram showing a conceptual configuration of a modification around the NN applied primary temperature controller in the air conditioning control device of the present embodiment.

FIG. 11 is a block diagram showing a conceptual configuration of a modification around the NN applied primary temperature controller in the air conditioning control device of the present embodiment.

[Explanation of symbols]

 Reference Signs List 10 large space 12 air conditioner 12A blower 12B heating / cooling coil 13 transport path 14 air conditioning air supply duct 15 air conditioning exhaust duct 16 temperature sensor 20 valve 22 immediately after temperature sensor 26 secondary temperature controller 28 measuring device 30 input device 32 NN applied primary Temperature controller 32A NN primary controller 32B controller

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yoshiaki Higuchi 1-5-1, Otsuka, Inzai City, Chiba Prefecture Inside Takenaka Corporation Technical Research Institute (72) Inventor Yoshiharu Asano 500 Wakasato, Nagano City, Nagano Prefecture F in Shinshu University Terms (reference) 3L060 AA06 CC02 CC11 CC19 DD08 EE21

Claims (5)

[Claims] 1. An air conditioning control apparatus for controlling an air conditioner to output a fluid whose temperature has been adjusted from the air conditioner to the target space when the fluid is supplied from the air conditioner to the target space, wherein a temperature in the target space is detected. Means, a measuring means for measuring a heat source in the target space, a setting means for setting a temperature in the target space, a temperature detection value in the target space, a heat source measurement value in the target space and a value in the target space. Input means for inputting a temperature set value, output means for outputting a control value of the air conditioner, intermediate means for transmitting data from the input means to the output means, and the input means or the output means The means are constituted by a neural network model having a multilayer structure in which connection relations of a large number of neural circuit elements are modeled, and a temperature detected value in the target space, a heat source measured value in the target space, and a To output a control value of the air-conditioning equipment corresponding to the temperature set point in the object space, between the input means and the intermediate means,
An air-conditioning control device, comprising: a corresponding unit that determines a connection relationship of the neural circuit element between the intermediate unit and the output unit in advance. 2. The input means of the corresponding means, using a relationship data in which a correspondence relationship between the plurality of temperature detection values, the heat source measurement values and the temperature set values, and a large number of control values of the air conditioner is known in advance. The air-conditioning control device according to claim 1, wherein a connection relation of the neural circuit elements included in the means to the output means is constructed. 3. The air conditioning control device according to claim 1, wherein the measuring unit measures a total number of persons existing in the target space. 4. The air conditioning control device according to claim 1, wherein the measurement unit measures a distribution of persons existing in the target space. 5. An air conditioner further comprising detecting means for detecting an output temperature of the air conditioner, wherein the corresponding means further causes an input means to input a detected value of the output temperature of the air conditioner, and causes the output means to output a temperature in the target space. The air-conditioning control device according to any one of claims 1 to 4, wherein a time to reach a set value is predicted.