Air conditioning system and method for controlling air conditioning system

Abstract

This air conditioning system comprises: an air conditioner controller for controlling an air conditioner provided in an air conditioning target space; a heat source unit controller for controlling a heat source unit and an auxiliary unit that supply a heat source to the air conditioner; and an integrated controller for outputting a control command to the air conditioner controller and the heat source unit controller.

Description

Air conditioning system and method for controlling air conditioning system

[0001] This application claims priority to Japanese Patent Application No. 2024-012749, filed on January 31, 2024, the contents of which are incorporated herein by reference.

[0002] In server rooms (data centers), buildings, etc., air conditioning systems are used that include a heat source machine such as an air-cooled heat pump chiller or a turbo refrigerator, and an air conditioner that cools the air in the room using a heat medium supplied from the heat source machine. For example, Patent Document 1 describes a technique for controlling the operating rate of each air conditioner so as to minimize the total power consumption of each of the multiple air conditioners. Furthermore, Non-Patent Documents 1 and 2 describe a technique for controlling the number of operating heat source machines so as to maximize the COP of a heat source system consisting of multiple heat source machines.

[0003] JP 2011-257062 A

[0004] Nikaido, Satoshi, Ueda, Kenji, Togano, Yoshie, Matsuo, Minoru, and Tateishi, Hiroki, "Ene-Conductor: A Controller for Optimal Control of Heat Source Systems Consisting of Centrifugal Chillers," Mitsubishi Heavy Industries Technical Review, Vol. 51, No. 2 (2014). Kikuchi, Hironari, Miyajima, Yuji, Ito, Fumihiro, and Maeyama, Akira, "Optimal Control System for Heat Source Equipment Aiming at CO2 Reduction and Energy Conservation," Hitachi Review, Vol. 96, No. 12 (2014).

[0005] In conventional technology, air conditioners and heat source units are independently controlled to achieve optimal conditions without coordination. Generally, when cooling a server room, for example, increasing the room temperature setting (set temperature) can reduce the load on the air conditioner and reduce power consumption. Therefore, server room managers tend to set the set temperature to a value close to the upper limit of the range that satisfies environmental requirements. However, while setting a set temperature in this way can minimize the power consumption of the air conditioner, it can be difficult to minimize the power consumption of the heat source unit.

[0006] An object of the present disclosure is to provide an air conditioning system and a control method for an air conditioning system that can minimize the power consumption of the entire air conditioning system while satisfying environmental requirements.

[0007] According to one aspect of the present disclosure, an air conditioning system includes an air conditioner controller that controls air conditioners installed in a space to be air-conditioned, a heat source machine controller that controls heat source machines and auxiliary machines that supply heat sources to the air conditioners, and an integrated controller that outputs control commands to the air conditioner controller and the heat source machine controller, wherein the integrated controller includes a first prediction unit that predicts the future heat load of the air-conditioned space and a valid range of a set temperature that satisfies environmental requirements of the space to be air-conditioned based on a first model that uses a heat load measured in the space to be air-conditioned as an explanatory variable and a future heat load prediction value of the space to be air-conditioned and a valid range of a set temperature that satisfies environmental requirements of the space to be air-conditioned as objective variables, and a power consumption prediction value of the air conditioner for each set temperature predicted by the air conditioner controller and a previous prediction unit that predicts the future heat load of the space to be air-conditioned and a valid range of a set temperature that satisfies environmental requirements of the space to be air-conditioned based on a first model that uses a heat load measured in the space to be air-conditioned as an explanatory variable and a future heat load prediction value of the space to be air-conditioned and a valid range of a set temperature that satisfies environmental requirements of the space to be air-conditioned as objective variables, and and a command unit that outputs, as the control command, the set temperature that minimizes the sum of the predicted power consumption values ​​based on the predicted power consumption values ​​of the heat source unit and the auxiliary unit for each set temperature predicted by the heat source unit controller, wherein the air conditioner controller has a second prediction unit that predicts the power consumption of the air conditioner for each set temperature included in the valid range using a second model that has the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as an objective variable, and the heat source unit controller has a third prediction unit that predicts the power consumption of the heat source unit and the auxiliary unit for each set temperature included in the valid range using a third model that has the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as an objective variable.

[0008] According to one aspect of the present disclosure, a control method for an air conditioning system includes an air conditioner controller that controls an air conditioner provided in an air-conditioned space, a heat source controller that controls a heat source machine and an auxiliary machine that supply a heat source to the air conditioner, and an integrated controller that outputs a control command to the air conditioner controller and the heat source machine controller, the integrated controller predicting a future heat load and a valid range of a set temperature that satisfies an environmental requirement of the air-conditioned space based on a first model in which a heat load measured in the air-conditioned space is used as an explanatory variable and a future heat load prediction value of the air-conditioned space and a valid range of a set temperature that satisfies an environmental requirement of the air-conditioned space are used as objective variables; The method includes a step of outputting, as the control command, the set temperature that minimizes the total predicted power consumption values ​​based on the predicted power consumption value of the air conditioner and the predicted power consumption values ​​of the heat source machine and the auxiliary machine for each set temperature predicted by the heat source machine controller; a step in which the air conditioner controller predicts the power consumption of the air conditioner for each set temperature included in the valid range using a second model in which the valid range of the set temperature and the predicted heat load value are explanatory variables and the predicted power consumption value for each set temperature is an objective variable; and a step in which the heat source machine controller predicts the power consumption of the heat source machine and the auxiliary machine for each set temperature included in the valid range using a third model in which the valid range of the set temperature and the predicted heat load value are explanatory variables and the predicted power consumption value for each set temperature is an objective variable.

[0009] According to the above aspect, it is possible to minimize the power consumption of the entire air conditioning system while satisfying environmental requirements.

[0010] 1 is a schematic diagram showing the overall configuration of an air conditioning system according to a first embodiment. FIG. 2 is a block diagram showing the functional configuration of an integrated controller according to the first embodiment. FIG. 3 is a diagram showing an example of a first model according to the first embodiment. FIG. 4 is a block diagram showing the functional configuration of an air conditioner controller according to the first embodiment. FIG. 5 is a diagram showing examples of second and third models according to the first embodiment. FIG. 6 is a block diagram showing the functional configuration of a heat source unit and auxiliary unit controller according to the first embodiment. FIG. 7 is a sequence diagram showing an example of processing in an air conditioning system according to the first embodiment. FIG. 8 is a diagram showing an example of a predicted power consumption value according to the first embodiment. FIG. 9 is a flowchart showing an example of processing in an integrated controller according to the first embodiment. FIG. 10 is a flowchart showing an example of processing in an air conditioner controller and a heat source unit and auxiliary unit controller according to the first embodiment. FIG. 11 is a schematic diagram showing the overall configuration of an air conditioning system according to a second embodiment. FIG. 12 is a diagram showing an example of a first model according to the second embodiment. FIG. 13 is a schematic diagram showing the overall configuration of an air conditioning system according to a third embodiment. FIG. 14 is a block diagram showing the functional configuration of a server controller according to the third embodiment. FIG. 15 is a sequence diagram showing an example of processing in an air conditioning system according to the third embodiment. FIG. 16 is a schematic diagram showing the overall configuration of an air conditioning system according to a fourth embodiment. FIG. 17 is a block diagram showing the functional configuration of a power generation controller according to the fourth embodiment. FIG. 18 is a flowchart showing an example of processing in a power generation controller according to the fourth embodiment.

[0011] First Embodiment Hereinafter, a first embodiment will be described in detail with reference to FIGS.

[0012] (Overall Configuration of Air Conditioning System) Fig. 1 is a schematic diagram showing the overall configuration of an air conditioning system according to a first embodiment. In this embodiment, an example will be described in which an air conditioning system 100 performs air conditioning (cooling) of a server room as shown in Fig. 1. Note that in other embodiments, the air conditioning system 100 may perform air conditioning (heating and cooling) of a building, a factory, a warehouse, etc.

[0013] As shown in FIG. 1 , the air conditioning system 100 includes an air conditioner 2, a heat source unit 3, an auxiliary unit 4, an integrated controller 10, an air conditioner controller 20, and a heat source unit / auxiliary unit controller 30.

[0014] The air conditioner 2 is an indoor unit installed inside the server room, which is the space to be air-conditioned. The air conditioner 2 draws air from the server room, exchanges heat with the heat medium of the heat source unit 3, and discharges the air to the server room. Multiple air conditioners 2 may be installed depending on the size of the server room and the number of servers.

[0015] The heat source unit 3 supplies a heat source to the air conditioner 2. The heat source unit 3 is, for example, an air-cooled heat pump chiller, a turbo chiller, or the like. In this embodiment, an example will be described in which the heat source unit 3 is a turbo chiller that supplies chilled water as a heat source. Multiple heat source units 3 may be installed depending on the size of the server room and the number of servers.

[0016] In this embodiment, the auxiliary machine 4 is a chilled water pump 4A. The auxiliary machine 4 may also include a cooling water pump 4B or an outdoor unit 4C (cooling tower) of a turbo chiller. In other embodiments, when the heat source machine 3 is an air-cooled heat pump chiller, the outdoor unit 4C may be configured as part of the heat source machine 3.

[0017] The integrated controller 10 controls the operation of the entire air conditioning system 100 so as to reduce the overall power consumption of the air conditioner side (Range 1) and the heat source side (Range 2) of the air conditioning system 100 while satisfying the environmental requirements of the server room. When the air conditioning system 100 is used to cool a server room as in this embodiment, the environmental requirement is an upper limit value for room temperature and is specified by the manager of the server room. Note that in other embodiments, the air conditioning system 100 may be used for heating and cooling a building, etc., in which case the environmental requirement may be specified as a range from a lower limit value to an upper limit value for room temperature.

[0018] The air conditioner controller 20 controls the operation of the air conditioner 2 so that the room temperature in the server room approaches the set temperature instructed by the integrated controller 10 .

[0019] The heat source machine / auxiliary machine controller 30 controls the operation of the heat source machine 3 and the auxiliary machine 4 (chilled water pump 4A) based on the set temperature commanded by the integrated controller 10. In the following description, the heat source machine / auxiliary machine controller 30 will also be abbreviated as the "heat source machine controller."

[0020] (Functional Configuration of Integrated Controller) Fig. 2 is a block diagram showing the functional configuration of the integrated controller according to embodiment 1. As shown in Fig. 2, the integrated controller 10 includes a processor 11, a memory 12, a storage 13, and a communication interface 14.

[0021] The processor 11 operates in accordance with a predetermined program to function as a first acquisition unit 101, a first prediction unit 102, a command unit 103, and a first learning unit 104.

[0022] The first acquisition unit 101 acquires measurement values ​​from sensors in each part of the air conditioning system 100. For example, the first acquisition unit 101 acquires a measurement value of the temperature (room temperature Ta_act) [°C] of the server room from a temperature sensor T1 installed in the server room. The first acquisition unit 101 acquires a measurement value of the outside air temperature Tf from a temperature sensor T2 installed outside the server room (e.g., near the outdoor unit 4C). The first acquisition unit 101 may also acquire measurement values ​​of the power consumption E1 and E2 [kW] of each part from power meters WM1, WM2, and WM3 installed on power lines (not shown) that supply power to the air conditioner 2, the heat source unit 3, and the auxiliary equipment 4. The first acquisition unit 101 also acquires predicted power consumption values ​​and operating states of the air conditioner 2, the heat source unit 3, and the auxiliary equipment 4 from the controllers 20 and 30. Details of the predicted power consumption values ​​will be described later.

[0023] The first prediction unit 102 uses a first model M1 ( FIG. 3 ) to predict the future heat load of the server room and the valid range of the set temperature that satisfies the environmental requirements of the server room. As shown in FIG. 3 , the first model M1 is a prediction model that uses the room temperature Ta_act measured in the server room as an explanatory variable and the predicted future heat load value of the server room as a response variable. The predicted future heat load value is, for example, the room temperature [°C] in the server room n minutes from the current time. The valid range of the set temperature is the set range of the room temperature (the discharge temperature of the cooling air from the air conditioner 2) in the server room that is below the environmental requirements (upper limit of the room temperature). The first model M1 may be trained to output the discharge flow rate of the cooling air, the chilled water temperature and flow rate, the cooling water temperature and flow rate, etc. The first model M1 may also use the current outside air temperature Tf, the date and time, the weather forecast, etc. as explanatory variables. By increasing the number of explanatory variables, the first model M1 can improve the accuracy of predicting future room temperatures by taking into account the influence of the outside air temperature, time of day, season, weather, etc.

[0024] The command unit 103 outputs to each of the controllers 20, 30, as a control command, a set temperature Ta that minimizes the total predicted power consumption value based on the predicted power consumption value for each set temperature predicted by each of the controllers 20, 30.

[0025] The first learning unit 104 performs machine learning to create a first model M1 using learning data such as time series of measured values ​​of the room temperature Ta_act and the outside air temperature Tf, time series of the set temperature Ta commanded by the command unit 103, weather, etc. In the learning phase, the first learning unit 104 learns the first model M1 using a predetermined number of pieces of learning data in advance and records the data in the storage 13. In addition, in the operation phase of the integrated controller 10, the first learning unit 104 sequentially collects learning data and learns (updates) the first model M1.

[0026] The memory 12 has a memory area necessary for the operation of the processor 11 .

[0027] The storage 13 is a so-called auxiliary storage device, such as a hard disk drive (HDD), a solid state drive (SSD), etc. The storage 13 stores data that each part of the processor 11 acquires, generates, and references during processing.

[0028] The communication interface 14 is an interface for transmitting and receiving data, control signals, measurements, etc. between the controllers 20, 30, 40 and various sensors.

[0029] (Functional configuration of air conditioner controller) Fig. 4 is a block diagram showing the functional configuration of the air conditioner controller according to Embodiment 1. As shown in Fig. 4, the air conditioner controller 20 includes a processor 21, a memory 22, a storage 23, and a communication interface 24.

[0030] The processor 21 operates in accordance with a predetermined program to function as a second acquisition unit 201, a second prediction unit 202, an air conditioner control unit 203, and a second learning unit 204.

[0031] The second acquisition unit 201 acquires measurement values ​​from sensors in each part of the air conditioning system 100. For example, the second acquisition unit 201 acquires a measurement value of the temperature (room temperature Ta_act) [°C] of the server room from a temperature sensor T1 installed in the server room. The second acquisition unit 201 acquires a measurement value of the power consumption E1 [kW] of the air conditioner 2 from a power meter WM1 of the air conditioner 2. The second acquisition unit 201 acquires a measurement value of the outside air temperature Tf from a temperature sensor T2 outside the server room. Note that the second acquisition unit 201 may acquire these measurement values ​​via the integrated controller 10 rather than directly from the sensors. The second acquisition unit 201 also acquires control commands from the integrated controller 10. The control commands include a request for a power consumption prediction value, a command for the set temperature Ta of the server room, etc.

[0032] The second prediction unit 202 predicts the power consumption E1 and operating point of the air conditioner 2 using a second model M2 ( FIG. 5 ). As shown in FIG. 5 , the second model M2 is a prediction model that uses the set temperature range and the predicted future heat load value of the server room as explanatory variables and the predicted power consumption value of the air conditioner 2 for each set temperature as a response variable. The second model M2 may also use the measured values ​​of the current room temperature Ta_act and the outside air temperature Tf as explanatory variables. By increasing the number of explanatory variables, the second model M1 can more accurately predict the power consumption E1 by taking into account the difference between the current room temperature and the outside air temperature and the predicted future room temperature.

[0033] The set temperature range includes at least a portion of the temperature range that satisfies the environmental requirements. When an upper limit value for room temperature (e.g., 15°C) is specified in the environmental requirements, the second prediction unit 202 inputs a predetermined range of temperatures below the upper limit value (e.g., 10°C to 15°C) into the second model M2 as the set temperature range. If a lower limit value is also specified in the environmental requirements, the second prediction unit 202 inputs the temperatures from the lower limit value to the upper limit value into the second model M2 as the set temperature range. The set temperature range may be set by the server room manager, and the integrated controller 10 may output it to the air conditioner controller 20 together with the heat load prediction value.

[0034] The air conditioner control unit 203 optimizes the operating conditions (the number of operating air conditioners 2, the air flow rate of the operating air conditioners 2, etc.) based on the set temperature Ta commanded by the integrated controller 10 to minimize the power consumption of the air conditioners 2, and controls the air conditioners 2. The process of optimizing the operating conditions based on the set temperature is the same as existing processes (see, for example, Patent Document 1), so a detailed description will be omitted.

[0035] The second learning unit 204 performs machine learning to generate a second model M2 using learning data such as a time series of temperature measurement values ​​(room temperature Ta1, outside air temperature Tf), a time series of the set temperature Ta, and a time series of measurement values ​​of power consumption E1 of the air conditioner 2. In the learning phase, the second learning unit 204 learns the second model M2 using a predetermined number of pieces of learning data in advance and records the data in the storage 23. In addition, in the operation phase of the air conditioner controller 20, the second learning unit 204 sequentially collects learning data and learns (updates) the second model M2.

[0036] The memory 22 has a memory area necessary for the operation of the processor 21 .

[0037] The storage 23 is a so-called auxiliary storage device, such as a hard disk drive (HDD), a solid state drive (SSD), etc. The storage 23 stores data that each part of the processor 21 acquires, generates, and references during processing.

[0038] The communication interface 24 is an interface for transmitting and receiving data, control signals, measurements, etc. between the integrated controller 10 and various sensors, etc.

[0039] (Functional configuration of the heat source machine / auxiliary machine controller) Fig. 6 is a block diagram showing the functional configuration of the heat source machine controller according to Embodiment 1. As shown in Fig. 6, the heat source machine controller 30 includes a processor 31, a memory 32, a storage 33, and a communication interface 34.

[0040] The processor 31 operates according to a predetermined program to perform the functions of a third acquisition unit 301, a third prediction unit 302, a heat source unit / auxiliary unit control unit 303, and a third learning unit 304.

[0041] The third acquisition unit 301 acquires measurement values ​​of sensors in each part of the air conditioning system 100. For example, the third acquisition unit 301 acquires a measurement value of the temperature (room temperature Ta_act) [°C] of the server room from a temperature sensor T1 installed in the server room. The third acquisition unit 301 acquires a measurement value of the power consumption E2 [kW] of the heat source unit 3 and the auxiliary unit 4 from a power meter WM2 of the heat source unit 3. The third acquisition unit 301 acquires a measurement value of the outside air temperature Tf from a temperature sensor T2 outside the server room. Note that the third acquisition unit 301 may acquire these measurement values ​​via the integrated controller 10 rather than directly from the sensors. The third acquisition unit 301 also acquires control commands from the integrated controller 10. The control commands include a request for a predicted power consumption value, a command for the set temperature Ta of the server room, etc.

[0042] The third prediction unit 302 predicts the power consumption E2 of the heat source unit 3 and the auxiliary unit 4 using a third model M3 ( FIG. 5 ). As shown in FIG. 5 , the third model M3 is a prediction model similar to the second model M2, and uses the set temperature range and the predicted future heat load value of the server room as explanatory variables, and the predicted power consumption value of the heat source unit 3 for each set temperature as a response variable. The set temperature range input to the third model M3 is the same as that input to the second model M2.

[0043] The heat source unit / auxiliary unit control unit 303 optimizes the operating conditions (the number of operating heat source units 3, the rotation speed of the compressors of the operating heat source units 3, the rotation speed of the auxiliary units 4, etc.) based on the set temperature Ta commanded by the integrated controller 10 so as to minimize the power consumption of the heat source units 3 and the auxiliary units 4, and controls the heat source units 3 and the auxiliary units 4. The process of optimizing the operating conditions based on the set temperature is the same as existing processes (see, for example, Non-Patent Documents 1 and 2), so a detailed description will be omitted. In the following description, the heat source unit / auxiliary unit control unit 303 will also be abbreviated and referred to as the "heat source unit control unit."

[0044] The third learning unit 304 performs machine learning to generate a third model M3 using, as learning data, a time series of temperature measurement values ​​(room temperature Ta1, outside air temperature Tf), a time series of the set temperature Ta, and a time series of measurement values ​​of power consumption E2 of the heat source unit 3 and the auxiliary unit 4. In the learning phase, the third learning unit 304 learns the third model M3 using a predetermined number of pieces of learning data in advance and records the data in the storage 33. In addition, in the operation phase of the heat source unit controller 30, the third learning unit 304 sequentially collects learning data and learns (updates) the third model M3.

[0045] The memory 32 has a memory area necessary for the operation of the processor 31 .

[0046] The storage 33 is a so-called auxiliary storage device, such as a hard disk drive (HDD), a solid state drive (SSD), etc. The storage 33 stores data that each part of the processor 31 acquires, generates, and references during processing.

[0047] The communication interface 34 is an interface for transmitting and receiving data, control signals, measurements, etc. between the integrated controller 10 and various sensors, etc.

[0048] (Processing of Air Conditioning System) FIG. 7 is a sequence diagram showing an example of processing of the air conditioning system according to the first embodiment. Here, the processing flow of the air conditioning system 100 will be described in detail with reference to FIG. 7. It is assumed that, at the time of FIG. 7, the learning phase of each of the controllers 10, 20, and 30 has been completed, and the learned models M1 to M3 have been recorded in the storage of each controller. Therefore, FIG. 7 shows the processing of the operation phase of each controller.

[0049] First, the first acquisition unit 101 of the integrated controller 10 acquires the measured values ​​(room temperature Ta_act in the server room, outdoor temperature Tf, etc.) at the current time from each sensor (step S101). At this time, the first acquisition unit 101 may also acquire weather forecast data from an external server (not shown).

[0050] After acquiring the measured values, the first prediction unit 102 of the integrated controller 10 uses the first model M1 to predict the future heat load and the range of the future set temperature (steps S102 and S103). Specifically, the first prediction unit 102 reads the first model M1 from the storage 13. The first prediction unit 102 inputs the measured values ​​(room temperature Ta_act in the server room, outdoor temperature Tf) acquired in step S101 at the current time into the read first model M1, and obtains the future heat load prediction value and the range of the future set temperature as outputs (steps S102 and S103). The first prediction unit 102 may also input the current date and time and weather forecast data into the first model M1. The heat load prediction value is, for example, the predicted room temperature [°C] in the server room n minutes from now. The range of the set temperature is, for example, the range of the set temperature (cooled air discharge temperature) of the air conditioner 2 that satisfies the environmental requirements of the server room n minutes from now (keeping the room temperature below the upper limit value).

[0051] In addition, the command unit 103 of the integrated controller 10 requests each controller 20, 30 to predict the power consumption of the air conditioner 2, the heat source unit 3, and the chilled water pump 4A (auxiliary equipment) at each set temperature included in the valid range of the set temperature of the air conditioner 2 based on the thermal load prediction value (step S104).

[0052] Upon receiving a request from the integrated controller 10, the second prediction unit 202 of the air conditioner controller 20 optimizes the operating conditions to minimize the power consumption of the air conditioner 2 at each set temperature (step S105). The second prediction unit 202 inputs the predicted heat load value obtained from the integrated controller 10 and the valid range of the set temperature obtained from the integrated controller 10 into the second model M2 read from the storage 23, and obtains a predicted value of the power consumption E1 of the air conditioner 2 for each set temperature as an output. For example, as shown in FIG. 8A, assume that the environmental requirement specifies that the room temperature in the server room be 15°C or less. Also, assume that the valid range of the set temperature (the valid range of the operating conditions of the air conditioner 2) is 10°C or higher and 15°C or lower. The second prediction unit 202 uses the second model M2 to predict the power consumption E1 of the air conditioner 2 when the set temperature is between 10°C and 15°C. FIG. 8A is a graph showing an example of the predicted value of the power consumption E1 of the air conditioner 2 versus the set temperature. The second prediction unit 202 also outputs the predicted power consumption values ​​of the air conditioners 2 for each set temperature and the operating conditions at that time to the integrated controller 10 (step S106).

[0053] Upon receiving a request from the integrated controller 10, the third prediction unit 302 of the heat source unit controller 30 optimizes the operating conditions to minimize the power consumption of the heat source units 3 and the auxiliary units 4 at each set temperature (step S107). For example, the third prediction unit 302 inputs the thermal load prediction value obtained from the integrated controller 10 and the valid range of the set temperature into the third model M3 read from the storage 33, and obtains as output a predicted value of the power consumption E2 of the heat source units 3 and the auxiliary units 4 for each set temperature. For example, in FIG. 8B, the dashed-line graph is a graph with the power consumption E2 on the vertical axis and the operating conditions of the heat source units 3 and the auxiliary units 4 on the horizontal axis. This graph shows the change in power consumption E2 when the operating conditions of the heat source units 3 (number of operating units, compressor rotation speed) are changed at a certain set temperature Ta. The solid-line graph in FIG. 8B is a graph with the power consumption E2 on the vertical axis and the set temperature Ta on the horizontal axis. The solid line graph connects the optimal points (operating points where power consumption E2 is minimum) of the dashed line graphs corresponding to each set temperature Ta, and represents the transition of power consumption E2 when the heat source unit 3 is controlled most efficiently at each set temperature Ta. As shown in the solid line graph, the power consumption characteristics of the heat source unit 3 and the auxiliary units 4 for the set temperature Ta differ from the power consumption characteristics of the air conditioner 2 ((a) of FIG. 8). The third model M3 learns these power consumption characteristics of the heat source unit 3 and the auxiliary units 4 and outputs predicted values ​​of power consumption of the heat source unit 3 and the auxiliary units 4 corresponding to each temperature within the valid range of the set temperature. In addition, the third prediction unit 302 outputs the predicted power consumption value of the heat source unit 3 for each set temperature and the operating conditions at that time to the integrated controller 10 (step S108).

[0054] Next, the command unit 103 of the integrated controller 10 acquires the predicted power consumption values ​​from each of the controllers 20 and 30 and selects the set temperature that minimizes the sum of the predicted power consumption values ​​(step S110). For example, as shown in FIG. 8C, when the set temperature is 13°C, the sum of the predicted power consumption E1 of the air conditioner 2 and the predicted power consumption E2 of the heat source unit 3 and the auxiliary unit 4 is minimized (optimal point P3). In this case, the command unit 103 selects the optimal point P3 = 13°C from the valid set temperature range (10°C to 15°C) as the set temperature for the server room. The command unit 103 outputs the selected set temperature (13°C) as a control command to each of the controllers 20 and 30 (step S110).

[0055] The air conditioner control unit 203 of the air conditioner controller 20 changes the operating conditions of the air conditioner 2 to the operating conditions obtained in step S105 that correspond to the set temperature commanded by the integrated controller 10 (13° C. in the example of FIG. 8) (step S111). Similarly, the heat source unit control unit 303 of the heat source unit controller 30 changes the operating conditions of the heat source unit 3 and the auxiliary unit 4 (chilled water pump 4A) to the operating conditions obtained in step S107 that correspond to the set temperature commanded by the integrated controller 10 (step S112).

[0056] Each controller repeatedly executes the series of processes shown in Fig. 7 at regular intervals. In this way, each component of the air conditioning system 100 can be operated in response to constantly changing room and outside temperatures, satisfying the environmental requirements of the server room, while minimizing the power consumption of the entire air conditioning system 100.

[0057] (Learning Process) Furthermore, each controller sequentially performs model learning and updating in parallel with the process shown in Fig. 7. Here, the learning process will be described with reference to Figs.

[0058] 9 is a flowchart showing an example of processing by the integrated controller according to the first embodiment. The integrated controller 10 performs the learning processing shown in FIG. 9 in parallel with the series of processing shown in FIG.

[0059] First, the first acquisition unit 101 of the integrated controller 10 acquires the measured value of the heat load (room temperature in the server room) at the current time (step S201).

[0060] Next, the first learning unit 104 of the integrated controller 10 reads, from the storage 13, a thermal load prediction value previously predicted by the first prediction unit 102 (step S202). Assume that the first prediction unit 102 predicted the thermal load n minutes from now in step S102 of Fig. 7. In this case, the first learning unit 104 reads the thermal load prediction value predicted n minutes before the current time.

[0061] The first learning unit 104 also learns the error between the actual thermal load (measured value) at the current time t and the predicted thermal load value at time t predicted n minutes ago, and updates the first model M1 (step S203). The first learning unit 104 updates the first model M1 by learning weights so as to minimize an evaluation function based on the error between the actual measured value and the predicted value, for example.

[0062] 9 shows an example in which the first model M1 learns a thermal load, but this is not limiting. In another embodiment, the first model M1 may further learn operating points such as the discharge temperature and flow rate of the cooling air of the air conditioner 2, the chilled water temperature and flow rate, and the cooling water temperature and flow rate. In this case, in steps S201 to S203, the first learning unit 104 learns errors between the measured values ​​and predicted values ​​of these operating points in addition to the thermal load, and updates the first model M1.

[0063] Fig. 10 is a flowchart showing an example of the processing of the air conditioner controller, heat source controller, and auxiliary controller according to the first embodiment. Each controller performs the learning processing shown in Fig. 10 in parallel with the series of processing in Fig. 7. Since the processing of each controller is the same, the air conditioner controller 20 will be described as an example.

[0064] First, the second acquisition unit 201 of the air conditioner controller 20 acquires the measurement of the power consumption E1 of the air conditioner 2 (step S211).

[0065] Next, the second learning unit 204 of the air conditioner controller 20 reads from the storage 23 the predicted power consumption value of the air conditioner 2 that was previously predicted by the second prediction unit 202 (step S212). Assume that the second prediction unit 202 predicted the power consumption n minutes from now in step S105 of Fig. 7. In this case, the second learning unit 204 reads the predicted power consumption value that was predicted n minutes before the current time.

[0066] The second learning unit 204 also learns the error between the actual power consumption E1 (measured value) of the air conditioner 2 at the current time t and the predicted power consumption value predicted n minutes ago at time t, and updates the second model M2 (step S213). The second learning unit 204 updates the second model M2 by learning weights so as to minimize an evaluation function based on the error between the actual measured value and the predicted value, for example. The third learning unit 304 of the heat source unit controller 30 similarly updates the third model M3 based on the error between the actual measured value and the predicted value of the power consumption E2 of the heat source unit 3 and the auxiliary unit 4 (chilled water pump 4A).

[0067] 10 shows an example in which the second model M2 and the third model M3 learn power consumption, but this is not limiting. In other embodiments, the second model M2 and the third model M3 may learn operating points such as the discharge temperature and flow rate of the cooling air, the chilled water temperature and flow rate, and the cooling water temperature and flow rate of the air conditioner 2. In this case, in steps S211 to S213, the second learning unit 204 and the third learning unit 304 respectively learn the errors between the measured values ​​and predicted values ​​of these operating points in addition to the power consumption, and update the second model M2 and the third model M3.

[0068] (Actions and Effects) As described above, the air conditioning system 100 includes the air conditioner controller 20, the heat source machine controller 30, and the integrated controller 10. The integrated controller 10 includes a first prediction unit 102 that predicts the future heat load (room temperature) of the server room and the valid range of the set temperature that satisfies the environmental requirements of the server room based on a first model M1 that uses the heat load (room temperature Ta_act) measured in the server room (air-conditioned space) as an explanatory variable and the predicted future heat load value of the server room and the valid range of the set temperature that satisfies the environmental requirements of the server room as objective variables, and a command unit 103 that outputs, as the control command, the set temperature that minimizes the sum of the predicted power consumption values ​​based on the predicted power consumption values ​​of the air conditioner 2 for each set temperature predicted by the air conditioner controller 20 and the predicted power consumption values ​​of the heat source machine 3 and the auxiliary machine 4 for each set temperature predicted by the heat source machine controller 30. The air conditioner controller 20 has a second prediction unit 202 that predicts the power consumption E1 of the air conditioner 2 for each set temperature using a second model M2 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a response variable.The heat source machine controller 30 has a third prediction unit 302 that predicts the power consumption E2 of the heat source machine 3 and the auxiliary machine 4 for each set temperature using a third model M3 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a response variable.

[0069] As described above, in conventional technology, server room managers tend to determine the set temperature for the server room (air conditioner) to minimize the power consumption of the air conditioner. However, as shown in Figures 8(a) and 8(b), the air conditioner, heat source unit, and auxiliary unit have different power consumption characteristics. Therefore, if the set temperature for the server room is set based on the optimal point for the air conditioner, it is difficult to minimize the power consumption of the heat source unit. In contrast, the air conditioning system 100 according to the present embodiment searches for and determines, from among multiple set temperatures, the set temperature that minimizes the total predicted power consumption of the air conditioner 2, heat source unit 3, and auxiliary unit 4, thereby minimizing the power consumption of the entire air conditioning system 100 while satisfying environmental requirements. Furthermore, compared to conventional technology, the set temperature for the server room can be appropriately set, thereby more effectively lowering the temperature of server components (CPU, GPU) and reducing the probability of server component failure. Furthermore, lowering the temperature of server components reduces the operating rate of the cooling fans equipped in the server, thereby reducing the power consumption of the server.

[0070] The heat load of the server room (air-conditioned space) is the room temperature of the server room, and the first prediction unit 102 of the integrated controller 10 predicts the future room temperature of the server room as a heat load prediction value.

[0071] In this way, the air conditioning system 100 can measure the room temperature Ta_act, which is the current heat load of the server room, with a simple configuration in which only the temperature sensor T1 is provided in the server room.

[0072] Second Embodiment A second embodiment will be described below with reference to Figures 11 and 12. Note that, among the configurations of the second embodiment, the same configurations as those of the first embodiment will be described using the same reference numerals as those of the first embodiment.

[0073] Fig. 11 is a schematic diagram showing the overall configuration of an air conditioning system according to a second embodiment. As shown in Fig. 11, in this embodiment, a plurality of temperature sensors T3 are further provided in the server room. Each temperature sensor T3 is provided adjacent to each of a plurality of servers 5, and measures the temperatures (CPU temperature, GPU temperature) Tp1, Tp2, Tp3, ... of each server 5.

[0074] 7, the first acquisition unit 101 of the integrated controller 10 acquires the measured values ​​of temperatures Tp1, Tp2, Tp3, ... of each server 5 at the current time from the temperature sensor T3. In step S102 of FIG. 7, the first prediction unit 102 further inputs the measured values ​​of temperatures Tp1, Tp2, Tp3, ... of each server 5 into the first model M1 and obtains a future heat load predicted value as an output. Note that, as shown in FIG. 12, the first model M1 in this embodiment outputs the predicted temperature [°C] of each server 5 n minutes from now as the heat load predicted value.

[0075] In this embodiment, the environmental requirement is the maximum allowable temperature of each server, and the set temperature is the temperature of each server. The maximum allowable temperature is, for example, a value defined as a specification of the server (CPU, GPU). In step S105 of FIG. 7, the second prediction unit 202 of the air conditioner controller 20 inputs a temperature within a predetermined range below the maximum allowable temperature (e.g., 40°C to 70°C) as the set temperature range into the second model M2. The processing in step S107 of FIG. 7 is similar.

[0076] In this way, the air conditioning system 100 can minimize the power consumption of the entire air conditioning system 100 while preventing each server 5 from exceeding its maximum allowable temperature. Furthermore, by monitoring and cooling the temperature of each server 5 in this manner, it is possible to cool the server components more effectively than in the first embodiment, thereby further reducing the probability of server (CPU, GPU) failure. Furthermore, by cooling the server components, the operating rate of the server's cooling fan can be reduced, improving the effect of reducing the server's power consumption.

[0077] <Third Embodiment> A third embodiment will be described below with reference to Figures 13 to 15. Note that, among the configurations of the third embodiment, the same configurations as those of the above-mentioned embodiments will be described using the same reference numerals as those of the above-mentioned embodiments.

[0078] 13 is a schematic diagram showing the overall configuration of an air conditioning system according to Embodiment 3. As shown in FIG. 13, the air conditioning system 100 according to this embodiment further includes a server controller 50.

[0079] Fig. 14 is a block diagram showing the functional configuration of a server controller according to the third embodiment. As shown in Fig. 14, the server controller 50 includes a processor 51, a memory 52, a storage 53, and a communication interface 54.

[0080] The processor 51 operates in accordance with a predetermined program to function as a fifth acquisition unit 501 and an allocation unit 502 .

[0081] The fifth acquisition unit 501 acquires a job request from a higher-level device to the server 5 .

[0082] The allocation unit 502 allocates jobs to each server 5 so that the processing volume of each server 5 is equalized.

[0083] The memory 52 has a memory area necessary for the operation of the processor 51 .

[0084] The storage 53 is a so-called auxiliary storage device, such as a hard disk drive (HDD), a solid state drive (SSD), etc. The storage 53 stores data that each part of the processor 51 acquires, generates, and references during processing.

[0085] The communication interface 54 is an interface for transmitting and receiving data, control signals, measurements, etc. between the integrated controller 10, higher-level devices, various sensors, etc.

[0086] In addition, when there is a server whose server temperature margin is below a threshold, the command unit 103 of the integrated controller 10 commands the air conditioner controller 20 to lower the set temperature of the air conditioner that is to air-condition that server below the set temperatures of the other air conditioners.

[0087] The air conditioner control unit 203 of the air conditioner controller 20 performs control to lower the set temperature of the designated air conditioner.

[0088] Fig. 15 is a sequence diagram showing an example of processing of the air conditioning system according to the third embodiment. Here, the processing flow of the integrated controller 10, the air conditioner controllers 20, and the server controller 50 according to this embodiment will be described with reference to Fig. 15 .

[0089] First, the fifth acquisition unit 501 of the server controller 50 acquires a job request from a higher-level device (step S301). Then, the allocation unit 502 of the server controller 50 changes the allocation of jobs to each server 5 to equalize the processing load of each server 5. This allows the server controller 50 to prevent the load from concentrating on some servers 5 and causing an increase in server temperature.

[0090] In parallel with the processing of the server controller 50, the first acquisition unit 101 of the integrated controller 10 acquires the measured values ​​of temperatures Tp1, Tp2, Tp3, . . . of each server 5 at the current time from the temperature sensor T3 (step S303).

[0091] The command unit 103 of the integrated controller 10 also determines whether there is a server whose server temperature margin is below a threshold (step S304). The margin is the difference between the maximum allowable temperature and the current server temperature (maximum allowable temperature - current server temperature). If there is a server 5 whose margin is below a threshold (e.g., 5°C) (step S304), the command unit 103 extracts the air conditioner 2 that is to air-condition this server 5 as the target air conditioner. Note that the storage 13 of the integrated controller 10 pre-stores information associating which air conditioner 2 and which server 5 are to air-condition, and the command unit 103 references this information to extract the target air conditioner. The command unit 103 also changes the target air conditioner's set temperature so that it is lower than the set temperatures of the other air conditioners 2 (step S305), and outputs the target air conditioner's identification information (air conditioner ID) and the changed set temperature as a control command to the air conditioner controller 20 (step S306).

[0092] The air conditioner control unit 203 of the air conditioner controller 20 performs control to optimize the operating conditions of the target air conditioner specified by the integrated controller 10 based on the changed set temperature (step S307).

[0093] The air conditioning system 100 repeatedly executes the series of processes shown in FIG. 15 at regular intervals. In this manner, the air conditioning system 100, through the processing of the server controller 50, can suppress unevenness in the processing load of the servers 5 and prevent the temperature of each server 5 from exceeding the maximum allowable temperature. Furthermore, even if the processing load is concentrated on some servers 5 and the temperature rises despite the server controller 50 attempting a temperature leveling process, the integrated controller 10 can change the set temperatures of some of the air conditioners 2 to cool only those servers 5. This prevents the temperature of each server 5 from exceeding the maximum allowable temperature and prevents the power consumption of the entire air conditioning system 100 from increasing.

[0094] <Fourth embodiment> A fourth embodiment will be described below with reference to Figures 16 to 18. Note that, among the configurations of the fourth embodiment, the same configurations as those of the above-mentioned embodiments will be described using the same reference numerals as those of the above-mentioned embodiments.

[0095] Fig. 16 is a schematic diagram showing the overall configuration of an air conditioning system according to the fourth embodiment. As shown in Fig. 16, the air conditioning system 100 according to this embodiment further includes a power generation facility 6 and a power generation controller 60.

[0096] The power generation facility 6 is, for example, a solar power generation facility or the like, and supplies the generated power to the air conditioning system 100. The power generation facility 6 has a storage battery 6A, and when the generated power (supply amount) is greater than the power consumption (demand amount) of the air conditioning system 100, the surplus power may be charged into the storage battery 6A. Furthermore, when the generated power is less than the power consumption of the air conditioning system 100, the shortfall may be covered by discharging power from the storage battery 6A, or by purchasing power from the power grid.

[0097] 17 is a block diagram showing the functional configuration of a power generation controller according to the fourth embodiment. As shown in FIG. 17, the power generation controller 60 includes a processor 61, a memory 62, a storage 63, and a communication interface 64.

[0098] The processor 61 operates in accordance with a predetermined program to function as a supply and demand prediction unit 601 and a power generation control unit 602 .

[0099] The supply and demand prediction unit 601 predicts the amount of power demand of the air conditioning system 100 and the amount of power supply of the power generation facility 6 .

[0100] The power generation control unit 602 adjusts the amount of power generated by the power generation facility 6 based on the amount of power demand and the amount of power supply.

[0101] The memory 62 has a memory area necessary for the operation of the processor 61 .

[0102] The storage 63 is a so-called auxiliary storage device, such as a hard disk drive (HDD), a solid state drive (SSD), etc. The storage 63 stores data that each part of the processor 61 acquires, generates, and references during processing.

[0103] The communication interface 64 is an interface for transmitting and receiving data, control signals, measurements, etc. between the integrated controller 10 and various sensors, etc.

[0104] 18 is a flowchart showing an example of processing by the power generation controller according to the fourth embodiment. Here, the flow of processing by the power generation controller 60 will be described with reference to FIG.

[0105] First, the supply and demand prediction unit 601 of the power generation controller 60 predicts the power demand of the air conditioning system 100 and the power supply of the power generation facility 6 (step S401). For example, the supply and demand prediction unit 601 obtains from the integrated controller 10 the sum of the predicted values ​​of power consumption of each unit for the set temperature, and sets this as the predicted value of the power demand n minutes from now. The supply and demand prediction unit 601 also predicts the amount of power generated (power supply) of the power generation facility 6 n minutes from now based on statistical data and the like accumulated in advance.

[0106] Next, if the predicted power supply amount is greater than the power demand amount (step S402; YES), the power generation control unit 602 of the power generation controller 60 determines that the generated power is surplus. In this case, the power generation control unit 602 performs control to reduce the generated power by the surplus or to charge the storage battery (step S403). For example, if the storage battery 6A can store the surplus, the power generation control unit 602 outputs a control command to the power generation facility 6 to store the surplus in the storage battery 6A. Furthermore, if the storage amount of the storage battery 6A is large and some or all of the surplus cannot be stored in the storage battery 6A, the power generation control unit 602 may output a control command to the power generation facility 6 to reduce the generated power. In this way, the power generation control unit 602 can suppress excessive power generation by the power generation facility 6. Furthermore, storing surplus power in the storage battery 6A can be used as a preparation for a time when the power demand is high.

[0107] Furthermore, if the predicted power supply amount is less than the power demand amount (step S402; NO and step S404; YES), the power generation control unit 602 performs control to increase the power generation amount of the power generation equipment 6, control to discharge the shortage from the storage battery 6A, control to purchase power from the power grid, or control to suppress server jobs (step S405). For example, if the storage amount of power in the storage battery 6A is sufficient, the power generation control unit 602 outputs a control command to the power generation equipment 6 to discharge the shortage from the storage battery 6A and supply it to the air conditioning system 100. Furthermore, if the storage amount of power in the storage battery 6A is not enough to cover the shortage or if the power generation amount of the power generation equipment 6 can be increased (i.e., the power supply amount < the available power generation amount), the power generation control unit 602 may output a control command to the power generation equipment 6 to increase the power generation amount. If the storage amount of power in the storage battery 6A is not sufficient and the power generation amount of the power generation equipment cannot be increased any further, the power generation equipment 6 requests the server controller 50 via the integrated controller 10 to suppress server jobs. In this case, the command unit 103 of the integrated controller 10 receives a request from the power generation controller 60 and commands the server controller 50 to suppress job submission. Upon receiving the command to suppress job submission, the server controller 50 restricts the acceptance of new jobs from the host device. Furthermore, if the power generation control unit 602 attempts to suppress job submission but the power demand exceeds the power supply, the power generation control unit 602 may purchase power from the power grid to make up for the shortfall. Note that the power generation control unit 602 may first attempt to suppress job submission, and if the power demand remains excessive, perform other control, such as increasing the amount of power generated. The administrator may arbitrarily change the priority of these controls. In this way, when excessive power demand is predicted in the air conditioning system 100, the power generation control unit 602 can suppress job submission to the server 5 and reduce power demand. Furthermore, the power demand of the air conditioning system 100 can be met by increasing the amount of power generated by the power generation facility 6, discharging power from the storage battery 6A, or purchasing power.

[0108] If the predicted power demand and power supply match (step S403; NO and step S404; NO), the power generation control unit 602 ends the process without taking any action.

[0109] The power generation controller 60 repeatedly executes the series of processes shown in FIG. 18 at regular intervals.

[0110] Fifth Embodiment A fifth embodiment will be described below with reference to Fig. 19. Note that, among the configurations of the fifth embodiment, the same configurations as those of the above-described embodiments will be described using the same reference numerals as those of the above-described embodiments.

[0111] FIG. 19 is a schematic diagram showing the overall configuration of an air conditioning system according to a fifth embodiment. FIG. 19 illustrates the flow of data, such as control commands and measurement values, in the air conditioning system 100. As shown in FIG. 19 , in the integrated controller 10 according to this embodiment, the first model M1 includes a server indoor environment optimization AI (M1a) and an air conditioning system operating point optimization AI (M1b). Like the first model M1 in the above-described embodiment, the server indoor environment optimization AI (M1a) inputs the current indoor environment (room temperature) and outdoor environment (outdoor temperature) and predicts and outputs the future heat load state of the server indoor environment and the acceptable ranges for the discharge temperature (set temperature) and flow rate of the cooled air of the air conditioner 2. The air conditioning system operating point optimization AI (M1b) predicts and outputs the optimal operating point of the air conditioning system that satisfies the acceptable ranges for the discharge temperature and flow rate of the cooled air. The optimal operating point includes, for example, the cooled air discharge temperature and flow rate, the coolant inlet temperature and flow rate, and the chilled water outlet temperature and flow rate. The air conditioner control unit 203 of the air conditioner controller 20 outputs a control signal to the air conditioner 2 to control the fan rotation speed based on the optimal operating point (cooled air discharge temperature and flow rate) output by the air conditioning system operating point optimization AI (M1b). The heat source machine controller 30 outputs control signals to each device, such as the cooling water pump rotation speed, cooling tower fan rotation speed, the number of operating heat source machines 3 and load distribution, and chilled water pump rotation speed, based on the optimal operating points (cooled water temperature and flow rate, chilled water temperature and flow rate) output by the air conditioning system operating point optimization AI (M1b). Even in this aspect, the same effects as in the above-described embodiment can be obtained.

[0112] <Other Embodiments> Although one embodiment has been described in detail above with reference to the drawings, the specific configuration is not limited to the above, and various design modifications are possible. That is, in other embodiments, the order of the above-described processes may be changed as appropriate. Furthermore, some processes may be executed in parallel.

[0113] <Additional Notes> The air conditioning system and the control method for the air conditioning system described in the above-described embodiment can be understood, for example, as follows.

[0114] (1) According to a first aspect, an air conditioning system 100 includes an air conditioner controller 20 that controls an air conditioner 2 provided in an air-conditioned space, a heat source unit controller 30 that controls a heat source unit 3 and an auxiliary unit 4 that supply a heat source to the air conditioner 2, and an integrated controller 10 that outputs control commands to the air conditioner controller 20 and the heat source unit controller 30. The integrated controller 10 includes a first prediction unit 102 that predicts the future heat load of the air-conditioned space and the valid range of the set temperature based on a first model M1 that uses the heat load measured in the air-conditioned space as an explanatory variable and the future heat load predicted value of the air-conditioned space and the valid range of the set temperature that satisfies the environmental requirements of the air-conditioned space as objective variables, and a first prediction unit 103 that predicts the consumption of the air conditioner 2 for each set temperature predicted by the air conditioner controller 20. The air conditioner controller 20 has a second prediction unit 202 that predicts the power consumption of the air conditioner 2 for each set temperature using a second model M2 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature included in the valid range as objective variables, and the heat source machine controller 30 has a third prediction unit 302 that predicts the power consumption of the heat source machine 3 and the auxiliary machine 4 for each set temperature using a third model M3 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature included in the valid range as objective variables.

[0115] The air conditioning system 100 searches for and determines, from among multiple set temperatures, the set temperature that minimizes the sum E1 + E2 of the predicted power consumption values ​​of the air conditioner 2, the heat source unit 3, and the auxiliary units 4, thereby minimizing the power consumption of the entire air conditioning system 100 while satisfying environmental requirements. Furthermore, compared to conventional technologies, the system can appropriately cool the server room, thereby more effectively lowering the temperature of server components (CPU, GPU) and reducing the probability of server component failure. Furthermore, lowering the temperature of server components reduces the operating rate of the cooling fans equipped in the servers, thereby reducing the power consumption of the servers.

[0116] (2) According to the second aspect, in the air conditioning system 100 relating to the first aspect, the heat load is the room temperature of the space to be air-conditioned, and the first prediction unit 102 of the integrated controller 10 predicts the future room temperature of the space to be air-conditioned as a heat load prediction value.

[0117] In this way, the air conditioning system 100 can measure the room temperature Ta_act, which is the current heat load of the server room, with a simple configuration in which only the temperature sensor T1 is provided in the server room.

[0118] (3) According to the third aspect, in the air conditioning system 100 relating to the first aspect, multiple servers 5 are installed in the air-conditioned space, the heat load of the air-conditioned space is the server temperature of each server 5, and the first prediction unit 102 of the integrated controller 10 predicts the future server temperature as a heat load prediction value.

[0119] In this way, the air conditioning system 100 can minimize the power consumption of the entire air conditioning system 100 while preventing each server 5 from exceeding its maximum allowable temperature. Furthermore, by monitoring and cooling the temperature of each server 5 in this manner, it is possible to cool the server components more effectively than in the first embodiment, thereby further reducing the probability of server (CPU, GPU) failure. Furthermore, by cooling the server components, the operating rate of the server's cooling fan can be reduced, improving the effect of reducing the server's power consumption.

[0120] (4) According to a fourth aspect, the air conditioning system 100 according to the third aspect further includes a server controller 50 that assigns jobs to the servers 5 so that the processing volume of each of the multiple servers 5 is equalized. When there is a server 5 whose server temperature margin is less than a threshold, the command unit 103 of the integrated controller 10 commands the air conditioner controller 20 to lower the set temperature of the air conditioner 2 that is to air-condition that server 5 below the set temperatures of the other air conditioners 2.

[0121] In this way, the air conditioning system 100 can suppress unevenness in the processing load of the servers 5 through the processing of the server controller 50, and can prevent the temperature of each server 5 from exceeding the maximum allowable temperature. Furthermore, even if the processing load is concentrated on some servers 5 and the temperature rises despite the server controller 50 attempting a temperature leveling process, the integrated controller 10 can change the set temperatures of some of the air conditioners 2 to cool only those servers 5. This makes it possible to prevent an increase in the power consumption of the entire air conditioning system 100 while preventing the temperature of each server 5 from exceeding the maximum allowable temperature.

[0122] (5) According to a fifth aspect, the air conditioning system 100 according to any one of the first to fourth aspects further includes a power generation controller 60 that controls the power generation equipment 6 that supplies generated power to the air conditioning system 100, and the power generation controller 60 has a supply and demand prediction unit 601 that predicts the power demand of the air conditioning system 100 and the power supply amount of the power generation equipment 6, and a power generation control unit 602 that adjusts the power generation amount of the power generation equipment 6 based on the power demand and the power supply amount.

[0123] In this way, the air conditioning system 100 can suppress excessive power generation by the power generation equipment 6.

[0124] (6) According to a sixth aspect, the air conditioning system 100 according to the fifth aspect further includes a server controller 50 that assigns jobs to the servers 5 so that the processing volume of each of the multiple servers 5 installed in the air-conditioned space is equalized. When the power demand is greater than the power supply, the power generation control unit 602 of the power generation controller 60 requests the server controller 50 to suppress job submission through the integrated controller 10. When the server controller 50 is requested to suppress job submission, it restricts the acceptance of new jobs.

[0125] In this way, when excessive demand for electricity is predicted, the air conditioning system 100 can reduce the demand for electricity by suppressing job submission to the server 5.

[0126] (7) According to a seventh aspect, a control method for an air conditioning system 100 is a control method for an air conditioning system 100 including an air conditioner controller 20 that controls an air conditioner 2 provided in an air-conditioned space, a heat source machine controller 30 that controls a heat source machine 3 and an auxiliary machine 4 that supply a heat source to the air conditioner 2, and an integrated controller 10 that outputs a control command to the air conditioner controller 20 and the heat source machine controller 30, the control method including the steps of: the integrated controller 10 predicting a future heat load of the air-conditioned space and a valid range of a set temperature that satisfies the set temperature range based on a first model M1 that uses a heat load measured in the air-conditioned space as an explanatory variable, and a future heat load predicted value of the air-conditioned space and a valid range of a set temperature that satisfies the environmental requirements of the air-conditioned space as objective variables; The method includes a step of outputting, as a control command, the set temperature that minimizes the total predicted power consumption values ​​based on the predicted power consumption values ​​of the air conditioner 2 for each set temperature measured and the predicted power consumption values ​​of the heat source unit 3 and the auxiliary unit 4 for each set temperature predicted by the heat source unit controller 30; a step in which the air conditioner controller 20 predicts the power consumption of the air conditioner 2 for each set temperature included in the valid range using a second model M2 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a target variable; and a step in which the heat source unit controller 30 predicts the power consumption of the heat source unit 3 and the auxiliary unit 4 for each set temperature included in the valid range using a third model M3 that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a target variable.

[0127] According to the above aspect, it is possible to minimize the power consumption of the entire air conditioning system while satisfying environmental requirements.

[0128] 100 Air conditioning system 2 Air conditioner 3 Heat source unit 4 Auxiliary unit 4A Chilled water pump 4B Cooling water pump 4C Outdoor unit 5 Server 6A Storage battery 10 Integrated controller 101 First acquisition unit 102 First prediction unit 103 Command unit 104 First learning unit 20 Air conditioner controller 201 Second acquisition unit 202 Second prediction unit 203 Air conditioner control unit 204 Second learning unit 30 Heat source unit controller 301 Third acquisition unit 302 Third prediction unit 303 Heat source unit control unit 304 Third learning unit 50 Server controller 501 Fifth acquisition unit 502 Allocation unit 60 Power generation controller 601 Supply and demand prediction unit 602 Power generation control unit

Claims

1. An air conditioning system comprising: an air conditioner controller that controls air conditioners installed in a space to be air-conditioned; a heat source controller that controls heat source machines and auxiliary machines that supply heat sources to the air conditioners; and an integrated controller that outputs control commands to the air conditioner controller and the heat source machine controller, wherein the integrated controller has: a first prediction unit that predicts the future heat load of the air-conditioned space and the valid range of a set temperature that satisfies the environmental requirements of the air-conditioned space based on a first model that uses the heat load measured in the space to be air-conditioned as an explanatory variable and the predicted future heat load value of the space to be air-conditioned and the valid range of a set temperature that satisfies the environmental requirements of the space to be air-conditioned as objective variables; and a command unit that outputs, as the control command, the set temperature that minimizes the sum of the predicted power consumption values, based on the predicted power consumption values of the air conditioners for each set temperature predicted by the air conditioner controller and the predicted power consumption values of the heat source machines and auxiliary machines for each set temperature predicted by the heat source machine controller, the air conditioner controller has a second prediction unit that predicts the power consumption of the air conditioner for each set temperature included in the valid range using a second model that uses the valid range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a response variable; and the heat source machine controller has a third prediction unit that predicts the power consumption of the heat source machine and the auxiliary machine for each set temperature included in the valid range using a third model that uses the predicted range of the set temperature and the predicted heat load value as explanatory variables and the predicted power consumption value for each set temperature as a response variable.

2. The air conditioning system of claim 1, wherein the heat load is the room temperature of the air-conditioned space, and the first prediction unit of the integrated controller predicts the future room temperature of the air-conditioned space as the heat load prediction value.

3. The air conditioning system of claim 1, wherein a plurality of servers are installed in the air-conditioned space, the heat load of the air-conditioned space is the server temperature of each of the servers, and the first prediction unit of the integrated controller predicts the future server temperatures as the heat load prediction value.

4. The air conditioning system of claim 3, further comprising a server controller that allocates jobs to the servers so that the processing volume of each of the multiple servers is equalized, and wherein the command unit of the integrated controller commands the air conditioner controller to lower the set temperature of the air conditioner that is to cool the server to a temperature lower than the set temperatures of the other air conditioners when there is a server whose server temperature margin is below a threshold.

5. An air conditioning system as described in any one of claims 1 to 4, further comprising a power generation controller that controls a power generation facility that supplies generated power to the air conditioning system, the power generation controller having: a supply and demand prediction unit that predicts the power demand of the air conditioning system and the power supply amount of the power generation facility; and a power generation control unit that adjusts the power generation amount of the power generation facility based on the power demand amount and the power supply amount.

6. The air conditioning system of claim 5, further comprising a server controller that allocates jobs to the servers so that the processing volume of each of the multiple servers installed in the air-conditioned space is equalized, wherein the power generation control unit of the power generation controller requests the server controller through the integrated controller to suppress job submission when the power demand is greater than the power supply, and the server controller restricts the acceptance of new jobs when the request to suppress job submission is received.

7. A control method for an air conditioning system comprising an air conditioner controller that controls air conditioners installed in a space to be air-conditioned; a heat source controller that controls heat source machines and auxiliary machines that supply heat sources to the air conditioners; and an integrated controller that outputs control commands to the air conditioner controller and the heat source machine controller, comprising: a step in which the integrated controller predicts the future heat load of the air-conditioned space and the valid range of a set temperature that satisfies the environmental requirements of the air-conditioned space based on a first model in which the heat load measured in the space to be air-conditioned is used as an explanatory variable and the future heat load predicted value of the air-conditioned space and the valid range of a set temperature that satisfies the environmental requirements of the space to be air-conditioned are used as objective variables; and a step in which the integrated controller outputs, as the control command, the set temperature that minimizes the total of the predicted power consumption values based on the predicted power consumption values of the air conditioners for each set temperature predicted by the air conditioner controller and the predicted power consumption values of the heat source machines and auxiliary machines for each set temperature predicted by the heat source machine controller; A control method for an air conditioning system, comprising: a step in which the air conditioner controller predicts the power consumption of the air conditioner for each set temperature included in the valid range using a second model in which the valid range of the set temperature and the predicted heat load value are explanatory variables and the predicted power consumption value for each set temperature is a response variable; and a step in which the heat source machine controller predicts the power consumption of the heat source machine and the auxiliary machine for each set temperature included in the valid range using a third model in which the valid range of the set temperature and the predicted heat load value are explanatory variables and the predicted power consumption value for each set temperature is a response variable.

Images