JPH08210686A - Air-conditioning system - Google Patents

Abstract

PURPOSE: To carry out an efficient control of an air-conditioning machine by a method wherein the control objective of the air-conditioning machine is determined based on the atmospheric data of outside air, time data, an estimated internal load and an estimated air-conditioning thermal load, the amount of control is operated based on the control objective and the air-conditioning machine is controlled based on the operated amount of control. CONSTITUTION: A control objective determining means 18 determines the control objective of an air-conditioning machine 23 based on the atmospheric data of outside air, outputted from an outside air atmosphere detecting means 20, time data, outputted from a time information detecting means 22, an internal load, estimated by an internal load estimating means, and an air-conditioning thermal load, estimated by an air-conditioning thermal load estimating means 14. A control amount operating means 19 operates the amount of control of the air-conditioning machine 23 based on the determined control objective while an air-conditioning machine control means 21 controls the air-conditioning machine 23 based on the operated amount of control. The internal load of a room to be air-conditioned can thereby be grasped, whereby the air- conditioning machine 23 can be controlled efficiently.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an air conditioning system,
In particular, the present invention relates to an air conditioning system that estimates an internal load and controls an air conditioner.

[0002]

2. Description of the Related Art Conventionally, in this type of air conditioning system, the prediction of the heat load is divided into the outside air load due to the outside air and the internal load due to the internal heating elements such as OA equipment, certifying equipment, and people in the room to be air-conditioned. Instead, they predicted the total amount of heat the air conditioner would remove. Also, regarding the internal heat generation amount for estimating the internal load, conventionally, a sensor such as an infrared camera is installed indoors for prediction, and information such as the day of the week and work schedules and the degree of use of the internal heating element are used. Was predicted from the relationship.

[0003]

However, the method using the infrared camera is disadvantageous in that the cost is high because the sensor is expensive, and the method using the day of the week or the work schedule has a problem in relation to the internal load. The drawback is that it takes months to understand. The present invention has been made in view of such circumstances, and an object thereof is to provide an air conditioning system capable of estimating the internal load of an air-conditioned room by a simple method and efficiently controlling the air conditioner.

[0004]

In order to achieve the above object, the present invention provides an air conditioner for air-conditioning an air-conditioned room, an outside air environment detecting means for detecting outside air environment data outside the air-conditioned room, and a time period. Time information detecting means for detecting data, air conditioning heat for predicting the air conditioning heat load of the air conditioner based on the outside air environment data output from the outside air environment detecting means and the time data output from the time information detecting means The air-conditioned room based on the load prediction means, the outside air environment data output from the outside air environment detection means, the time data output from the time information detection means, and the air conditioning heat load predicted by the air conditioning heat load prediction means. Internal load estimating means for estimating an internal load of the external air environment data, external air environment data detected by the external air environment detecting means, time data detected by the time information detecting means, A control target determining unit that determines a control target of the air conditioner based on an internal load estimated by the partial load estimating unit and an air conditioning heat load estimated by the air conditioning heat load estimating unit; and the control target determining unit. From control amount calculation means for calculating the control amount of the air conditioner based on the determined control target, and air conditioner control means for controlling the air conditioner based on the control amount calculated by the control target calculation means. It is characterized by being formed.

In order to achieve the above object, an air conditioner for air conditioning the air-conditioned room, an outside air environment detecting means for detecting outside air environment data outside the air-conditioned room, and a time information detecting means for detecting time data. And a device removal heat amount detection means for detecting a device removal heat amount which is the heat amount to be actually removed from the air-conditioned room, an outside air environment data output from the outside air environment detection means and a time output from the time information detection means. Internal load estimation means for estimating the internal load of the air-conditioned room based on the data and the device removal heat amount detected by the device removal heat amount detecting means, the outside air environment data detected by the outside air environment detecting means, and the time. The time data detected by the information detecting means, the internal load estimated by the internal load estimating means, and the device removal heat quantity detecting means are detected. Control target deciding means for deciding a control target of the air conditioner on the basis of the amount of heat removed and installed; And an air conditioner control unit that controls the air conditioner based on the control amount calculated by the control target calculation unit.

[0006]

According to the first aspect of the present invention, the air conditioning system is based on the outside air environment data output from the outside air environment detecting means and the time data output from the time information detecting means by the air conditioning heat load predicting means. Predicting the air conditioning heat load, the internal load estimating means, the air conditioning heat load predicted by the air conditioning heat load predicting means, the outside air environment data output from the outside air environment detecting means, and the time output from the time information detecting means. The internal load of the air-conditioned room is estimated based on the data. Then, by the control target determining means, the outside air environment data output from the outside air environment detecting means, the time data output from the time information detecting means, the internal load estimated by the internal load estimating means, and the air conditioning heat load. The control target of the air conditioner that air-conditions the air-conditioned room is determined based on the air conditioning heat load predicted by the prediction means, and the control amount calculation means calculates the control amount of the air conditioner based on the determined control target. . The air conditioner control means controls the air conditioner based on the control amount calculated by the control amount calculation means. In this way, by grasping the internal load of the air-conditioned room, it becomes possible to efficiently control the air conditioner.

According to the second aspect of the present invention, in the air conditioning system, the internal load estimating means detects the device removal heat amount detected by the device removal heat amount detecting means and the outside air environment data and time information output from the outside air environment detecting means. The internal load of the air-conditioned room is estimated based on the time data output from the detection means. Then, by the control target determining means, the outside air environment data output from the outside air environment detecting means, the time data output from the time information detecting means, the internal load estimated by the internal load estimating means, and the device removal heat amount. The control target of the air conditioner that air-conditions the air-conditioned room is determined based on the device removal heat amount detected by the detection unit, and the control amount calculation unit calculates the control amount of the air conditioner based on the determined control target. . The air conditioner control means controls the air conditioner based on the control amount calculated by the control amount calculation means.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of an air conditioning system according to the present invention will be described below in detail with reference to the accompanying drawings. FIG. 1 shows a block diagram of a first embodiment of an air conditioning system according to the present invention. As shown in the figure, the air conditioning system 10
Is mainly the detection means 12 and the air conditioning heat load prediction means 1
4, internal load estimation means 16, control target determination means 18, control amount calculation means 19, air conditioner control means 21, and air conditioner 23
Consists of

The detecting means 12 comprises an outside air environment detecting means 20 and a time information detecting means 22, as shown in detail in FIG. The outside air environment detecting means 20
Is composed of an outside air temperature detecting means 26 for detecting an outside air temperature, an outside air humidity detecting means 28 for detecting an outside air humidity, and an insolation amount detecting means 30 for detecting an insolation amount, and the outside air temperature information detected by each detecting means. The outside air humidity information and the solar radiation amount information are used as outside air environment information, and the air conditioning heat load prediction means 1
4 and the internal load estimating means 16.

The time information detecting means 22 detects the time and the day of the week, and uses the time information and the day of the week information as the time information, like the outside air environment detecting means 20, as the air conditioning heat load predicting means 14 and the internal load estimating means 16. Is output to. The air conditioning heat load predicting means 14 includes a storage means 32 and a predicting means 3.
4 and learning means 36, the air conditioning heat load is predicted based on the outside air environment information and time information output from the detection means 12. Here, the air conditioning heat load is the amount of heat to be removed or the amount of heat to be supplied in order to maintain the temperature and humidity of the air-conditioned room.

The storage means 32 stores a neural network in which the relationship in which the outside air temperature, the outside air humidity, the solar radiation amount, the time and the day of the week are input and the air conditioning heat load is output is learned in advance. FIG. 3 shows a functional explanatory diagram of the neural network, and FIG. 4 shows the configuration of the neural network stored in the storage means 32 of the device removal heat amount detecting means 14.

In the neural network, when performing a network operation (forward operation), the input value is x ₁ ,
x ₂ , ..., X _n , internal coefficients (weights) w ₁ , w ₂ , ...
, W _n , the internal output value X of the accumulation processing unit 60 is And the output value Y from the neuro unit 62 is Y = 1 / (1 + exp (-AX)) where A is an arbitrary constant.

A plurality of neuro units 62 are connected in a plurality of layers to form a neural network, and the input / output relationships of the connected neuro units are as described above. When the neural network is modified (backward calculation), the internal coefficient (weight) is adjusted so that the accuracy is improved by using the back-propagation method based on the prediction accuracy of the predicted value.

The prediction means 34 uses the neural network stored in the storage means 32 to detect the detection means 12.
The predicted value of the device removal heat quantity is calculated by inputting the outside air environment information (outside air temperature, outside air humidity, solar radiation amount) and time information (time, day of the week) output from The learning unit 36 determines whether to newly learn the neural network stored in the storage unit 32 based on the prediction accuracy of the predicted value of the device removal heat amount predicted by the prediction unit 34, and determines that it is better to learn. In this case, the internal coefficient (weight) of the 5-input 1-output neural network including the three layers of the input layer 64, the intermediate layer 66, and the output layer 68 is corrected by the back propagation method based on the prediction accuracy.

The internal load estimating means 16 is a storage means 3.
8, the calculation means 40 and the estimation means 42, the outside air environment information and time information output from the detection means 12 and the air conditioning heat load predicted by the air conditioning heat load predicting means 14 into the internal load (OA in the air-conditioned room). Equipment, certification equipment, people, etc.). The storage means 38 stores the outside air temperature,
The neural network in which the relationship in which the outside air humidity, the solar radiation amount, the time, and the day of the week are input and the device removal heat amount is output is learned in advance is stored. Here, the device removal heat quantity is
In the heat load calculation, the heat load to be actually removed from the air-conditioned room is the air-conditioner heat load plus the heat storage load stored in the wall of the building. That is, when the room temperature is kept constant by air conditioning, the air conditioning heat load and the device removal heat amount are the same, but, for example, when the air conditioning is started up or when the air conditioning is intermittently operated, the walls of the building are exposed to sunlight or the like. The heat stored in
After a certain period of time, the room temperature of the air-conditioned room changes. Therefore, in order to perform accurate air conditioning control, it is necessary to know the device removal heat amount.

The neural network stored in the storage means 38 is constructed using a dynamic heat load calculation program 44, as shown in FIG. The dynamic heat load calculation program 44 has time data 46 in which time data and day of the week data are stored, standard weather data 48 in which outside air temperature data, outside air humidity data, and solar radiation amount data are stored. Enter the building conditions (building shape, material, etc.) of the building to be built and the indoor load conditions (internal heating elements in the room such as people, equipment, and lighting, quantity and amount of heat generated by individual heating elements). The amount of heat removed from the equipment for one year can be calculated every hour.

In this embodiment, the building condition and the indoor load condition are the design maximum load (internal load condition 1), 1/2 the design maximum load (internal load condition 2), and 1 / the design maximum load. In the case of 4 (internal load condition 3), three patterns are set and input to the dynamic heat load calculation program 44, and the device removal heat amount for each is calculated. As a result, from the dynamic heat load calculation program 44, the device removal heat amount 54 when the device removal heat amount is the design maximum load, the device removal heat amount 56 when the internal load is ½ of the design maximum load, and The device removal heat amount 58 when the internal load is ¼ of the designed maximum load is calculated.

The amount of heat removed by the apparatus 54 calculated by dividing the internal load conditions (building condition and indoor load condition) into three patterns,
Device removal heat amounts 54, 5
A neural network having three patterns is constructed by learning the relationships with the outputs of 6 and 58, the outside air temperature, the outside air humidity, the amount of solar radiation, the time, and the day of the week.
To memorize.

In the calculation means 40, the storage means 38
The outside air environment information and the time information output from the detection means 12 are input to each of the three patterns of neural networks stored in the above, and three patterns of device removal heat amount are calculated. The estimation unit 42 compares the air conditioning heat load predicted by the air conditioning heat load prediction unit 14 with the three patterns of device removal heat amounts calculated by the calculation unit, and is predicted by the air conditioning heat load prediction unit 14. The air-conditioning heat load is the device removal heat quantity 54, 5 of the three patterns calculated by the calculation means.
6, 58, 56, 58 of device removal heat amount calculated using a neural network of which internal load condition
And the internal load is estimated (which of the internal load conditions 1 to 3 is estimated) is estimated.

The control target determining means 18 includes the device removal heat quantity predicted by the air conditioning heat load predicting means 14, the internal load estimated by the internal load estimating means 16, the outside air environment data output from the outside air environment detecting means 20, and The control target of the air conditioner is determined from the time data output from the time information detecting means. The control amount calculation means 19 calculates the control amount of the air conditioner based on the control target determined by the control target determination means 18.

The air conditioner control means 21 controls the air conditioner based on the control amount calculated by the control amount calculation determining means 19. The air conditioner 23 air-conditions the air-conditioned room based on the operation signal output from the air conditioner control means 19. The operation of the air conditioning system 10 of the first embodiment configured as described above is as follows.

First, when the air conditioning system 10 is started, the outside air environment detecting means 20 and the time information detecting means 22
The data of the outside temperature, the outside humidity, the amount of solar radiation, the time of day, and the day of the week are detected. The air-conditioning heat load predicting means 14 predicts the air-conditioning heat load by inputting the outside air temperature, the outside air humidity, the amount of solar radiation, the time of day, and the day of the week output from the outside air environment detecting means 20 and the time information detecting means 22.

This prediction method is based on the air conditioning heat load prediction means 14
Of the outside air environment detecting means 20 and time to the neural network in which the relationship between the outside air temperature, the outside air humidity, the amount of solar radiation, the time and the day of the week and the air conditioning heat load stored in the storage means 32 is learned in advance. By inputting the data of the outside air temperature, the outside air humidity, the amount of solar radiation, the time of day, and the day of the week, which are output from the information detecting means 22, the network operation is performed,
Obtain the predicted value of the air conditioning heat load.

Then, the internal load estimating means 16 outputs the outside air temperature, the outside air humidity, the amount of solar radiation, the time, the day of the week output from the outside air environment detecting means 20 and the time information detecting means 22 and the air conditioning heat load predicting means. The air conditioning heat load predicted in 14 is input to estimate the internal load. In this estimation method, first, the calculation means 40 detects the outside air environment in a neural network in which the relationship between the outside air temperature, the outside air humidity, the amount of solar radiation, the time, and the day of the week and the device removal heat amount stored in the storage means 38 is learned in advance. The device removal heat quantity is calculated by inputting data of the outside air temperature, the outside air humidity, the amount of solar radiation, the time of day, and the day of the week, which are output from the means 20 and the time information detecting means 22, and performing a network operation.

Here, when the internal load condition is the design maximum load (internal load condition 1), the storage unit 38 uses the dynamic heat load calculation program as described above, and is half the design maximum load. Since the neural network constructed by the three patterns of the case (internal load condition 2) and the case of 1/4 of the design maximum load (internal load condition 3) is stored, the outside air temperature, the outside air humidity, When data of solar radiation amount, time and day of the week are input and the network calculation is performed by the calculation unit 40, different heat removal amount of the device is calculated from the calculation unit 40 depending on the internal load condition. That is, the device removal heat amount 54 when the device removal heat amount is the design maximum load, and the device removal heat amount 56 and the internal load when the internal load is 1/2 of the design maximum load is 1 of the design maximum load.
The device removal heat amount 58 when / 4 is calculated.

The estimating means 42 has three patterns of heat removal amounts 54, 56, 58 of the apparatus calculated by the calculating means 40,
The air-conditioning heat load predicting means 14 compares the air-conditioning heat load predicted by the air-conditioning heat load predicting means 14 with the device removal heat amount calculated by the neural network of which internal load condition among the set internal load conditions 1 to 3. It is judged whether or not it is the closest to the air conditioning heat load predicted in 14. As a result, the internal load is estimated from the set internal load conditions 1 to 3.
(It is assumed that the internal load is the internal load condition of the neural network that has calculated the device removal heat amount that most approximates the air conditioning heat load predicted by the air conditioning heat load prediction unit 14.) The internal load estimation unit 16 estimates the internal load. The internal load and the air conditioning heat load predicted by the air conditioning heat load predicting means 14 are input to the control target determining means 18, and the control target determining means 18 is supplied.
Then, the control target of the air conditioner is set on the basis of the internal load and the air conditioning heat load, and further, the outside air temperature, the outside air humidity, the amount of solar radiation, the time and the day of the week data output from the outside air environment detecting means 20 and the time information detecting means 22. decide. For example, when the weather load is large and the internal load is small in the summer, the control target is set to air-condition the perimeter zone. Further, if the device removal heat amount calculated by the calculation unit 40 of the internal load estimation unit 16 is further input to the data input to the control target determination unit 18 to set a control target in consideration of the heat storage load, It is possible to set a precise control target.

The control amount calculation means 19 calculates the control amount of the air conditioner 23 based on the control target determined by the control target determination means 18, and the air conditioner control means 21 calculates it by the control amount calculation setting means 19. An operation signal is output to the air conditioner 23 based on the control amount thus determined, and operation control is performed. In order to perform the above-mentioned operation every time step (1 hour),
The neural network stored in the storage means 32 of the air conditioning heat load predicting means 14 is updated so that the prediction accuracy is improved by the learning action, and more accurate air conditioning heat load prediction is possible.

As described above, according to the air conditioning system 10 of the first embodiment, it is possible to grasp the internal load of the air-conditioned room to be air-conditioned, and to efficiently control the air conditioner. Also,
Unlike the conventional forecasting method, it does not require expensive sensors, so the device cost can be reduced, and the schedule can be forecasted using a neural network. (Schedule forecasting can be done by inputting time and day of the week data to the neural network. By doing so, it becomes unnecessary to grasp the day of the week and the schedule.

FIG. 6 is a block diagram of the second embodiment of the air conditioning system according to the present invention. In the first embodiment, the air conditioning heat load used for estimating the internal load is predicted by the air conditioning heat load predicting means 14. In the second embodiment, the device removal heat amount detecting means 70 is used instead of the air conditioning heat load predicting means 14 to detect the device removal heat amount. The same members and the same devices as those in the first embodiment are designated by the same reference numerals and the description thereof will be omitted.

As shown in FIG. 6, the device removal heat amount detected by the device removal heat amount detecting means 70 is output to the internal load estimating means 16 and the control target determining means 18. In the internal load estimating means 16, similar to the first embodiment, the outside air temperature, the outside air humidity, the amount of solar radiation, the time of day, the data of the day of the week, and the amount of heat removed by the device are output from the outside air environment detecting means 20 and the time information detecting means 22. The device removal heat quantity predicted by the detection means 70 is input to estimate the internal load. The internal load estimated by the internal load estimating unit 16 and the device removing heat amount detected by the device removing heat amount detecting unit 70, and the outside air temperature and the outside air humidity output from the outside air environment detecting unit 20 and the time information detecting unit 22, The control target determining means 18 determines the control target of the air conditioner 23 based on the amount of solar radiation, the time, and the data of the day of the week. Then, the control amount calculation means 19 calculates the control amount of the air conditioner 23 based on the control target determined by the control target determination means 18, and the air conditioner control means 21.
Thus, an operation signal is output to the air conditioner 23 based on the control amount calculated by the control amount calculation determining means 19 to perform operation control.

As described above, also in the second embodiment, it is possible to grasp the internal load of the air-conditioned room in which the air conditioning is performed as in the first embodiment, and it is possible to efficiently control the air conditioner. In the present embodiment, the neural network used in the internal load estimating means 16 uses the dynamic heat load calculation program to set the internal load conditions in three patterns (design maximum load, 1 / design maximum load).
2, 1/4 of the maximum design load) was set and learned,
The setting pattern of the internal load condition is not limited to this. For example, the design maximum load, the design maximum load 9/1
The set patterns may be increased such as 0, ..., 2/10 of the design maximum load, and 1/10 of the design maximum load. This makes it possible to more accurately grasp the internal load. Further, as described above, the internal load may be predicted not by the level but by the numerical value.

Further, in this embodiment, the time information detecting means 22 detects the time and uses the information for each hour to predict the internal load and the heat removal amount of the apparatus. The daily load may be detected to predict the internal load and the amount of heat removed by the device. This makes it possible to plan the air conditioning for the next day. Further, in the present embodiment, the internal load conditions 1 to 3 are set for the load of all the buildings and the indoor load, but since the dynamic heat load calculation program can be calculated for each room, the above-mentioned operation is performed for each room. It is also possible to make an example neural network, calculate the amount of heat in each room from the temperature and flow rate of the suction port and the discharge port in each room, and estimate the internal load as in the above embodiment. Further, the above-described embodiment can be applied not only to each room but also to each floor and each block.
As a result, more efficient and comfortable air conditioning can be performed.

[0033]

As described above, according to the air conditioning system of the present invention, since the internal load can be grasped without using an expensive sensor as in the conventional case, efficient air conditioner control can be performed. In addition to being able to reduce costs, it is possible to reduce costs. Further, according to the air conditioning system of the present invention, the internal load can be grasped more quickly and the accuracy can be improved as compared with the conventional method of grasping the internal load from the schedule such as the day of the week or the work.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a first embodiment of an air conditioning system according to the present invention.

FIG. 2 is a configuration diagram of a first embodiment of an air conditioning system according to the present invention.

FIG. 3 is a functional explanatory diagram of a neural network.

FIG. 4 is a configuration diagram of a neural network used for device removal heat amount detecting means.

FIG. 5 is an explanatory diagram of a neural network used in the internal load estimating means.

FIG. 6 is a configuration diagram of a second embodiment of an air conditioning system according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 ... Air-conditioning system 12 ... Detection means 14 ... Air-conditioning heat load prediction means 16 ... Internal load estimation means 18 ... Control target determination means 19 ... Control amount calculation means 20 ... Outside air environment detection means 21 ... Air conditioner control means 22 ... Time information detection Means 23 ... Air conditioner 70 ... Device removal heat amount detecting means

Claims (9)

[Claims] 1. An air conditioner for air conditioning an air-conditioned room, an outside air environment detecting means for detecting outside air environment data outside the air-conditioned room, a time information detecting means for detecting time data, and an outside air environment detecting means. An air conditioning heat load predicting unit for predicting an air conditioning heat load of the air conditioner based on the outside air environment data output and the time data output from the time information detecting unit, and an outside air environment output from the outside air environment detecting unit Internal load estimation means for estimating the internal load of the air-conditioned room based on the data, the time data output from the time information detection means, and the air conditioning heat load predicted by the air conditioning heat load prediction means; and the outside air environment detection The outside air environment data detected by the means, the time data detected by the time information detecting means, the internal load estimated by the internal load estimating means, and Control target determining means for determining a control target of the air conditioner based on the air conditioning heat load predicted by the air conditioning heat load predicting means, and control of the air conditioner based on the control target determined by the control target determining means. An air conditioning system comprising: a control amount calculation means for calculating an amount; and an air conditioner control means for controlling the air conditioner based on the control amount calculated by the control target calculation means. 2. An air conditioner for air conditioning an air-conditioned room, an outside air environment detecting means for detecting outside air environment data outside the air-conditioned room, a time information detecting means for detecting time data, and an actual air conditioner from the air-conditioned room. Device removal heat amount detecting means for detecting device removal heat amount which is the amount of heat to be removed, outside air environment data output from the outside air environment detecting means, time data output from the time information detecting means, and the device removal heat amount detecting means Internal load estimation means for estimating the internal load of the air-conditioned room based on the amount of heat removed by the device, external air environment data detected by the external air environment detection means, and time detected by the time information detection means. Based on the data, the internal load estimated by the internal load estimating unit, and the device removal heat amount detected by the device removal heat amount detecting unit, Control target determining means for determining a control target of the controller, control amount calculating means for calculating a control amount of the air conditioner based on the control target determined by the control target determining means, and calculation by the control target calculating means An air conditioning system comprising: an air conditioner control unit that controls the air conditioner based on the controlled amount. 3. The air conditioning heat load predicting means has learned beforehand the relationship between the outside air environment data output from the environment information detecting means and the time data output from the time information detecting means, and the air conditioning heat load. Storage means in which a neural network is stored, and outside air environment data output from the outside air environment detection means and time data output from the time information detection means using the neural network stored in the storage means And a learning means for learning the neural network stored in the storage means based on the prediction accuracy of the predicted value of the air conditioning heat load calculated by the prediction means. The air conditioning system according to claim 1, comprising: 4. The apparatus, wherein the internal load estimating means is an outside air environment data output from the environment information detecting means and time data output from the time information detecting means, and an amount of heat to be actually removed from the air-conditioned room. Outputting from the outside air environment detecting means by using a storage means storing a plurality of neural networks in which the relationship with the amount of heat removed is learned in a plurality of patterns and a plurality of neural networks stored in the storage means. Calculating means for calculating the device removal heat quantity of a plurality of patterns by inputting the outside air environment data and the time data output from the time information detection means, the device removal heat quantity calculated by the calculation means and the air conditioning heat load. 2. An air conditioning system according to claim 1, further comprising: an estimating unit that estimates an internal load by comparing the air conditioning heat load estimated by the estimating unit. system. 5. The apparatus, wherein the internal load estimating means is an outside air environment data output from the environment information detecting means and time data output from the time information detecting means, and a heat quantity to be actually removed from the air-conditioned room. Outputting from the outside air environment detecting means by using a storage means storing a plurality of neural networks in which the relationship with the amount of heat removed is learned in a plurality of patterns and a plurality of neural networks stored in the storage means. Calculating means for calculating the device removal heat quantity of a plurality of patterns by inputting the outside air environment data and the time data output from the time information detection means, and the device removal heat quantity and the device removal heat quantity calculated by the calculation means 3. An estimation means for estimating an internal load by comparing with a device removal heat quantity output from the detection means. Air conditioning system. 6. The neural network stored in the storage means of the internal load estimating means sets a plurality of internal load conditions, inputs the set plurality of internal load conditions into a dynamic heat load calculation program, The device removal heat amount is calculated for each internal load condition, and the relationship between the calculated device removal heat amount and the outside air environment data output from the environment information detecting means and the time data output from the time information detecting means is neural-coded. The air conditioning system according to claim 4, wherein the network is trained. 7. The internal load conditions set and input in the dynamic thermal load calculation program are 3 in case of design maximum load, 1/2 of design maximum load and 1/4 of design maximum load. 7. Air conditioning system according to claim 6, characterized in that there are three internal load levels. 8. The outside air environment data detected by the outside air environment detecting means is an outside air temperature, an outside air humidity, and an amount of solar radiation, according to claim 1, 2, 3, 4, 5, 6 or 7. Air conditioning system. 9. The air conditioning system according to claim 1, wherein the time data detected by the time information detecting means is time and day of the week.