Tuning method, control device and tuning device

Abstract

To appropriately correct a parameter value which shows the magnitude of internal heat generation set for a prediction model in order to more accurately predict an air conditioning load in a target space by using a prediction model.SOLUTION: A tuning method performs tuning of a prediction model M1 that predicts an air conditioning load in a target space in a building. The prediction model M1 has a first parameter and a second parameter. The first parameter shows the magnitude of internal heat generation in the target space, and the second parameter is related to heat characteristics of the building. The tuning method includes a first step and a second step. In the first step, a value of the second parameter in the prediction model M1 for which a value of the first parameter has been set is corrected. In the second step, the value of the first parameter is corrected by using the prediction model M1 for which the corrected value of the second parameter has been set by performing the first step.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a tuning method, a control device, and a tuning device. [Background technology]

[0002] As disclosed in Patent Document 1 (Japanese Patent No. 7125644), there is a technique for predicting the air conditioning load in a target space using a prediction model in which a parameter value indicating the magnitude of internal heat generation is set. Summary of the Invention [Problem to be solved by the invention]

[0003] The magnitude of internal heat generation changes depending on the season. Therefore, in order to more accurately predict the air conditioning load in the target space using the prediction model, it is desirable to appropriately correct the value of the parameter indicating the magnitude of internal heat generation set in the prediction model. [Means for solving the problem]

[0004] The tuning method of the first aspect is performed by a tuning device. The tuning method tunes a prediction model. The prediction model predicts the air conditioning load in a target space within a building. The prediction model has a first parameter and a second parameter. The first parameter is a parameter that indicates the magnitude of internal heat generation in the target space. The second parameter is a parameter related to the thermal characteristics of the building. The tuning method includes a first step and a second step. The first step corrects the value of a second parameter in the prediction model in which the value of the first parameter is set, using an actual measurement value of the air conditioning load in the target space for a first period. The second step corrects the value of the first parameter set in the prediction model, using the actual measurement value of the air conditioning load in the target space for a second period and a predicted value of the air conditioning load for the second period predicted using the prediction model in which the value of the second parameter corrected by performing the first step is set.

[0005] In the tuning method of the first aspect, the prediction model has a first parameter and a second parameter. The first parameter is a parameter that indicates the magnitude of internal heat generation in the target space. The second parameter is a parameter related to the thermal characteristics of the building. The tuning method corrects the value of the first parameter in accordance with the correction of the value of the second parameter. As a result, the prediction model tuned by the tuning method can more accurately predict the air conditioning load in the target space.

[0006] A tuning method according to a second aspect is the tuning method according to the first aspect, in which the tuning device repeats a first step in a first period and a second step in a second period. The first step corrects the value of a second parameter in a prediction model in which the value of the first parameter corrected by performing the second step in a second period immediately before the first period is set.

[0007] The tuning method according to the second aspect can efficiently improve the accuracy of the prediction model by repeating the correction of the value of the first parameter and the correction of the value of the second parameter.

[0008] A tuning method according to a third aspect is the tuning method according to the first or second aspect, in which the first parameter has a value for each of a plurality of times including a first time. In a second step, a predicted value of the air conditioning load at the first time is calculated using a prediction model. In the second step, the value of the first parameter at the first time is corrected based on the difference between the actual measured value of the air conditioning load at the first time and the predicted value of the air conditioning load at the first time.

[0009] The tuning method of the third aspect corrects the value of the first parameter for each of a plurality of times, and thereby the prediction model tuned by the tuning method can predict the air conditioning load in the target space in more detail.

[0010] A tuning method according to a fourth aspect is the tuning method according to the third aspect, wherein the second step calculates a plurality of predicted values ​​of air conditioning loads at a first time in a second period. The second step calculates a representative value of the differences at the first time using a plurality of differences obtained by subtracting the corresponding actual measured values ​​of the air conditioning loads at the first time from the calculated predicted values ​​of the air conditioning loads at the first time. If the representative value is positive, the second step corrects the value of the first parameter at the first time to a smaller value. If the representative value is negative, the second step corrects the value of the first parameter at the first time to a larger value.

[0011] With this configuration, the tuning method according to the fourth aspect can correct the value of the first parameter without being biased towards the magnitude of internal heat generation on a particular day.

[0012] A tuning method according to a fifth aspect is the tuning method according to the fourth aspect, wherein the second step determines a correction amount for the value of the first parameter at the first time in accordance with the representative value.

[0013] A tuning method according to a sixth aspect is the tuning method according to the fourth aspect, in which the second period is one week.

[0014] With this configuration, the tuning method according to the sixth aspect can correct the value of the first parameter without being biased towards the magnitude of internal heat generation on a particular day of the week.

[0015] A tuning method according to a seventh aspect is the tuning method according to the first or second aspect, wherein the prediction model further includes a third parameter. The third parameter is a parameter related to the heat flow response of a wall of the building. The first step corrects the value of the third parameter using a plurality of patterns of values ​​that the third parameter can take. The plurality of patterns are determined in advance according to the magnitude of the heat capacity of the wall.

[0016] With this configuration, the tuning method according to the seventh aspect can efficiently correct the value of the third parameter.

[0017] A control device according to an eighth aspect includes a control unit. The control unit predicts an air conditioning load in a target space using a prediction model tuned by any one of the tuning methods according to the first to seventh aspects. The control unit determines a control target value for an indoor unit installed in the target space based on the predicted air conditioning load in the target space.

[0018] A tuning device of a ninth aspect tunes a prediction model. The prediction model predicts an air conditioning load in a target space within a building. The tuning device includes a control unit. The prediction model has a first parameter and a second parameter. The first parameter is a parameter indicating the magnitude of internal heat generation in the target space. The second parameter is a parameter related to the thermal characteristics of the building. The control unit corrects the value of the second parameter in the prediction model in which the value of the first parameter is set, using an actual measured value of the air conditioning load in the target space for a first period. The control unit corrects the value of the first parameter set in the prediction model, using the actual measured value of the air conditioning load in the target space for a second period and a predicted value of the air conditioning load for the second period predicted using the prediction model in which the corrected value of the second parameter is set.

[0019] In a tuning device of a ninth aspect, the prediction model has a first parameter and a second parameter. The first parameter is a parameter indicating the magnitude of internal heat generation in a target space. The second parameter is a parameter related to the thermal characteristics of the building. The tuning device corrects the value of the first parameter in accordance with the correction of the value of the second parameter. As a result, the prediction model tuned by the tuning device can more accurately predict the air conditioning load in the target space. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 2 is a schematic configuration diagram of a tuning device and an air conditioning device. [Figure 2] FIG. 2 is a functional block diagram of the air conditioning apparatus. [Figure 3] FIG. 2 is a functional block diagram of a tuning device. [Figure 4] As an example, this figure shows three patterns of values ​​that TFow,j (j=0 to 20) can take. [Figure 5] 10 is a flowchart illustrating an example of processing performed by the tuning device. DETAILED DESCRIPTION OF THE INVENTION

[0021] (1) Overall structure FIG. 1 is a schematic diagram of a tuning device 1 and an air conditioning device 2. As shown in FIG.

[0022] The tuning device 1 tunes a prediction model M1 that predicts the air conditioning load in a target space TS in a building BL.

[0023] The air conditioner 2 is configured with a vapor compression refrigeration cycle and provides air conditioning for one or more target spaces TS within the building BL. In this embodiment, the air conditioner 2 is a so-called multi-type air conditioning system for buildings. The air conditioner 2 may also be, for example, a central air conditioning system.

[0024] As shown in FIG. 1, a tuning device 1 and an air conditioner 2 are communicatively connected via a network NW such as the Internet.

[0025] (2) Detailed configuration (2-1) Air conditioning equipment As shown in Fig. 1, the air conditioner 2 mainly has a control device 40 and one or more refrigerant systems RS. Each refrigerant system RS has one outdoor unit 30 and one or more indoor units 20. The outdoor units 30 and indoor units 20 belonging to the same refrigerant system RS are connected by a liquid refrigerant communication pipe and a gas refrigerant communication pipe to form a refrigerant circuit. The control device 40 and the outdoor units 30 are connected to each other so that they can communicate with each other via a communication line 91. The outdoor units 30 and indoor units 20 belonging to the same refrigerant system RS are connected to each other so that they can communicate with each other via a communication line 90.

[0026] Each target space TS is installed with one or more indoor units 20. Also, each target space TS is installed with a solar radiation sensor 95. The solar radiation sensor 95 and the control device 40 are connected to each other so as to be able to communicate with each other via a network such as a wireless LAN.

[0027] Fig. 2 is a functional block diagram of the air conditioner 2. Fig. 2 shows, as representatives, one indoor unit 20 and one outdoor unit 30 that belong to the same refrigerant system RS. For simplicity, in this embodiment, the target space TS in which one indoor unit 20 is installed will be described.

[0028] (2-1-1) Indoor unit The indoor unit 20 is installed, for example, on the ceiling of the target space TS. The indoor unit 20 mainly has an indoor heat exchanger, an indoor fan, an indoor expansion valve 23, and an indoor control unit 29. The indoor unit 20 also has various sensors such as an indoor intake temperature sensor 61 and an indoor heat exchanger temperature sensor 62.

[0029] The indoor heat exchanger exchanges heat between the refrigerant flowing through it and the air in the target space. The indoor fan draws air from the target space into the indoor unit 20, exchanges heat with the refrigerant in the indoor heat exchanger, and supplies the air to the target space. The indoor fan is driven by the indoor fan motor 22m. The indoor expansion valve 23 is a mechanism for adjusting the pressure and flow rate of the refrigerant flowing through the refrigerant circuit. The indoor suction temperature sensor 61 measures the temperature of the air in the target space drawn in by the indoor unit 20. The indoor heat exchanger temperature sensor 62 measures the temperature of the refrigerant flowing through the indoor heat exchanger.

[0030] The indoor control unit 29 controls the operation of each component constituting the indoor unit 20. As shown in FIG. 2 , the indoor control unit 29 is communicatively connected to the indoor fan motor 22m and the indoor expansion valve 23. The indoor control unit 29 is also communicatively connected to various sensors, such as the indoor intake temperature sensor 61 and the indoor heat exchanger temperature sensor 62. The indoor control unit 29 has a control and arithmetic device and a storage device. The control and arithmetic device is a processor such as a CPU or a GPU. The storage device is a storage medium such as a RAM, a ROM, or a flash memory. The control and arithmetic device reads programs stored in the storage device and performs predetermined arithmetic processing in accordance with the programs, thereby controlling the operation of each component constituting the indoor unit 20. The control and arithmetic device can also write arithmetic results to the storage device and read information stored in the storage device in accordance with the programs. The indoor control unit 29 is configured to receive various signals transmitted from an operation remote control corresponding to the indoor unit 20. The indoor control unit 29 also exchanges various information, such as control signals, signals related to measurements by various sensors, and signals related to various settings, with the outdoor control unit 39 of the outdoor unit 30 via a communication line 90.

[0031] (2-1-2) Outdoor unit The outdoor unit 30 is installed, for example, on the roof of building BL. The outdoor unit 30 mainly has a compressor, a flow path switching valve 32, an outdoor heat exchanger, an outdoor expansion valve 34, an outdoor fan, and an outdoor control unit 39. The outdoor unit 30 also has various sensors such as an outdoor temperature sensor 66.

[0032] The compressor draws low-pressure refrigerant through the suction pipe, compresses the refrigerant using the compression mechanism, and discharges the compressed refrigerant through the discharge pipe. The compression mechanism of the compressor is driven by a compressor motor 31m. The flow path switching valve 32 switches the refrigerant flow path between a first state and a second state. During cooling operation, the flow path switching valve 32 sets the refrigerant flow path to the first state. At this time, the refrigerant discharged from the compressor flows through the refrigerant circuit in the following order: outdoor heat exchanger, outdoor expansion valve 34, indoor expansion valve 23, and indoor heat exchanger, before returning to the compressor. In the first state, the outdoor heat exchanger functions as a condenser, and the indoor heat exchanger functions as an evaporator. During heating operation, the flow path switching valve 32 sets the refrigerant flow path to the second state. At this time, the refrigerant discharged from the compressor flows through the refrigerant circuit in the following order: indoor heat exchanger, indoor expansion valve 23, outdoor expansion valve 34, and outdoor heat exchanger, before returning to the compressor. In the second state, the outdoor heat exchanger functions as an evaporator, and the indoor heat exchanger functions as a condenser. The outdoor heat exchanger exchanges heat between the refrigerant flowing through it and the outdoor air of the building BL. The outdoor expansion valve 34 is a mechanism for adjusting the pressure and flow rate of the refrigerant flowing through the refrigerant circuit. The outdoor fan supplies the outdoor air of the building BL to the outdoor heat exchanger. The outdoor fan is driven by an outdoor fan motor 36m. The outdoor temperature sensor 66 measures the temperature of the outdoor air of the building BL drawn in by the outdoor unit 30.

[0033] The outdoor control unit 39 controls the operation of each component constituting the outdoor unit 30. As shown in FIG. 2, the outdoor control unit 39 is communicatively connected to the compressor motor 31m, the flow path switching valve 32, the outdoor expansion valve 34, and the outdoor fan motor 36m. The outdoor control unit 39 is also communicatively connected to various sensors, such as the outdoor temperature sensor 66. The outdoor control unit 39 has a control and arithmetic device and a storage device. The control and arithmetic device is a processor such as a CPU or a GPU. The storage device is a storage medium such as a RAM, a ROM, or a flash memory. The control and arithmetic device reads programs stored in the storage device and performs predetermined arithmetic processing in accordance with the programs, thereby controlling the operation of each component constituting the outdoor unit 30. The control and arithmetic device can also write arithmetic results to the storage device and read information stored in the storage device in accordance with the programs. The outdoor control unit 39 exchanges various information, such as control signals, signals related to measurements by various sensors, and signals related to various settings, with the indoor control unit 29 of the indoor unit 20 via a communication line 90. Furthermore, the outdoor control unit 39 exchanges various types of information, such as control signals, signals relating to measurements by various sensors, and signals relating to various settings, with the control device 40 via a communication line 91.

[0034] (2-1-3) Control device The control device 40 centrally controls one or more refrigerant systems RS. The control device 40 is, for example, a device called an edge. The control device 40 is installed, for example, in a server room in the building BL. As shown in FIG. 2 , the control device 40 mainly includes a storage unit 41, a communication unit 44, and a control unit 49.

[0035] The storage unit 41 is a storage medium such as RAM, ROM, flash memory, etc. The storage unit 41 stores programs executed by the control unit 49, data necessary for executing the programs, etc. The communication unit 44 includes a network interface device for communicating with the tuning device 1 via the network NW, a network interface device for communicating with the outdoor unit 30 via the communication line 91, and a network interface device for communicating with the solar radiation sensor 95 via a network such as a wireless LAN.

[0036] The control unit 49 is a processor such as a CPU or GPU. The control unit 49 centrally controls one or more refrigerant systems RS by reading out programs stored in the storage unit 41 and performing predetermined arithmetic processing in accordance with the programs. The control unit 49 can also write arithmetic results to the storage unit 41 and read out information stored in the storage unit 41 in accordance with the programs. The control unit 49 exchanges various information, such as control signals, signals related to measurements by various sensors, and signals related to various settings, with the indoor control unit 29 of the indoor unit 20 and the outdoor control unit 39 of the outdoor unit 30 via a communication line 91. The control unit 49 also exchanges various information, such as control signals, signals related to measurements by various sensors, and signals related to various settings, with the tuning device 1 via a network NW.

[0037] The control unit 49 periodically acquires, as operating data D1, from the indoor control unit 29 of the indoor unit 20 and the outdoor control unit 39 of the outdoor unit 30, the airflow rate of the indoor fan (calculated from the rotation speed of the indoor fan motor 22m), the opening degree of the indoor expansion valve 23, the indoor intake temperature (measured value of the indoor intake temperature sensor 61), the indoor heat exchanger temperature (measured value of the indoor heat exchanger temperature sensor 62), the rotation speed of the compressor motor 31m, the opening degree of the outdoor expansion valve 34, the airflow rate of the outdoor fan (calculated from the rotation speed of the outdoor fan motor 36m), and the outdoor temperature (measured value of the outdoor temperature sensor 66). The control unit 49 also periodically calculates the air conditioning load (the heat exchange amount of the indoor heat exchanger) as operating data D1. The air conditioning load is calculated, for example, by inputting the airflow rate of the indoor fan, the indoor intake temperature, and the indoor heat exchanger temperature into a heat exchange function. The control unit 49 also periodically acquires a measurement value (amount of solar radiation) from the solar radiation sensor 95 as the operating data D1. The control unit 49 also periodically calculates an equivalent outdoor air temperature as the operating data D1. The equivalent outdoor air temperature is calculated from the outdoor temperature and amount of solar radiation. In this embodiment, the control unit 49 acquires operating data D1 at the top of the hour, such as 0:00, 1:00, etc., every hour. The control unit 49 stores the acquired operating data D1 in the memory unit 41.

[0038] The control unit 49 periodically transmits the acquired driving data D1 to the tuning device 1. In this embodiment, the control unit 49 transmits the acquired driving data D1 to the tuning device 1 every time it acquires driving data D1.

[0039] The control unit 49 acquires the prediction model M1 from the tuning device 1. In this embodiment, the control unit 49 acquires the prediction model M1 from the tuning device 1 every time the tuning device 1 corrects the prediction model M1. The control unit 49 stores the acquired prediction model M1 in the storage unit 41.

[0040] The control unit 49 predicts the air conditioning load in the target space TS using the prediction model M1 tuned by the tuning device 1. Specifically, the control unit 49 inputs the outdoor temperature, indoor temperature, equivalent outdoor air temperature, and solar radiation (items shown in Table 6, described later) at the time of the desired prediction, as well as past indoor temperatures, into the prediction model M1, and predicts the air conditioning load in the target space TS. The outdoor temperature, indoor temperature, equivalent outdoor air temperature, and solar radiation at the time of the desired prediction may be obtained, for example, from a weather forecast, or may be predicted using machine learning or the like from the outdoor temperature, indoor intake temperature, equivalent outdoor air temperature, and solar radiation in the operating data D1. The past indoor temperature is obtained, for example, from the indoor intake temperature in the operating data D1.

[0041] The control unit 49 determines the control target value of the indoor unit 20 installed in the target space TS based on the predicted air conditioning load in the target space TS.

[0042] For example, the control unit 49 calculates the indoor fan airflow rate and indoor heat exchanger temperature required to process the predicted air conditioning load in the target space TS. The indoor fan airflow rate and indoor heat exchanger temperature are calculated, for example, using machine learning or the like from the relationship between the air conditioning load, indoor fan airflow rate, and indoor heat exchanger temperature in the operating data D1. The control unit 49 then calculates the power consumption required to operate the air conditioner 2 based on the calculated indoor fan airflow rate and indoor heat exchanger temperature. The control unit 49 then determines the set temperature (control target value) for the indoor unit 20, taking into consideration the balance between the calculated power consumption and comfort in the target space TS.

[0043] When the control unit 49 receives an instruction to start cooling operation or heating operation from the operation remote control corresponding to the indoor unit 20, it switches the flow path switching valve 32 to the first state or the second state. Then, the control unit 49 adjusts the rotation speed of the compressor motor 31m, the opening degree of the outdoor expansion valve 34, the rotation speed of the outdoor fan motor 36m, the rotation speed of the indoor fan motor 22m, the opening degree of the indoor expansion valve 23, etc. so that the temperature (evaporation temperature or condensation temperature) of the refrigerant flowing through the indoor heat exchanger becomes a temperature corresponding to the determined set temperature.

[0044] For example, the control unit 49 determines the indoor heat exchange temperature (control target value) of the indoor unit 20 at the time to be predicted by inputting the indoor intake temperature (the indoor temperature at the time to be predicted that is input into the prediction model M1), the airflow rate of the indoor fan, and the air conditioning load at the time to be predicted into the heat exchange function. The airflow rate of the indoor fan at the time to be predicted is calculated using machine learning or the like from the relationship between the air conditioning load and the airflow rate of the indoor fan in the operating data D1, for example.

[0045] When the control unit 49 receives an instruction to start cooling operation or heating operation from the operation remote control corresponding to the indoor unit 20, it switches the flow path switching valve 32 to the first state or the second state. Then, the control unit 49 adjusts the rotation speed of the compressor motor 31m, the opening of the outdoor expansion valve 34, the rotation speed of the outdoor fan motor 36m, the rotation speed of the indoor fan motor 22m, the opening of the indoor expansion valve 23, etc. so that the temperature (evaporation temperature or condensation temperature) of the refrigerant flowing through the indoor heat exchanger becomes the determined indoor heat exchange temperature and so that the airflow rate of the indoor fan becomes the calculated airflow rate of the indoor fan.

[0046] For example, the control unit 49 determines the airflow rate (control target value) of the indoor fan of the indoor unit 20 at the time point to be predicted by inputting the indoor intake temperature (the indoor temperature at the time point to be predicted that is input into the prediction model M1), the indoor heat exchanger temperature, and the air conditioning load at the time point to be predicted into the heat exchange function. The indoor heat exchanger temperature at the time point to be predicted is calculated, for example, using machine learning or the like from the relationship between the air conditioning load and the indoor heat exchanger temperature in the operating data D1.

[0047] When the control unit 49 receives an instruction to start cooling operation or heating operation from the operation remote control corresponding to the indoor unit 20, it switches the flow path switching valve 32 to the first state or the second state. Then, the control unit 49 adjusts the rotation speed of the compressor motor 31m, the opening of the outdoor expansion valve 34, the rotation speed of the outdoor fan motor 36m, the rotation speed of the indoor fan motor 22m, the opening of the indoor expansion valve 23, etc. so that the temperature (evaporation temperature or condensation temperature) of the refrigerant flowing through the indoor heat exchanger becomes the calculated indoor heat exchange temperature and so that the airflow rate of the indoor fan becomes the determined indoor fan airflow rate.

[0048] (2-2) Tuning device The tuning device 1 is, for example, a computer installed on the cloud. FIG. 3 is a functional block diagram of the tuning device 1. As shown in FIG. 3, the tuning device 1 mainly includes a storage unit 11, a communication unit 14, and a control unit 19. In this embodiment, the functions of the tuning device 1 described below are realized by a single device. However, the functions of the tuning device 1 may also be realized in a distributed manner by multiple devices.

[0049] The storage unit 11 is a storage medium such as RAM, ROM, and flash memory. The storage unit 11 stores programs executed by the control unit 19, data necessary for executing the programs, etc. The communication unit 14 includes a network interface device for communicating with the control device 40 via the network NW.

[0050] (2-2-1) Control Unit The control unit 19 is a processor such as a CPU or a GPU. The control unit 19 reads and executes programs stored in the storage unit 11 to realize various functions of the tuning device 1. The control unit 19 can also write calculation results to the storage unit 11 and read information stored in the storage unit 11 according to the programs.

[0051] As shown in FIG. 3, the control unit 19 has, as functional blocks, an acquisition unit 191, a generation unit 192, a first correction unit 193, a second correction unit 194, and a transmission unit 195.

[0052] (2-2-1-1) Acquisition section The acquisition unit 191 periodically acquires the driving data D1 from the control device 40. In this embodiment, the acquisition unit 191 acquires the driving data D1 at the top of the hour every hour. The acquisition unit 191 stores the acquired driving data D1 in the storage unit 11.

[0053] (2-2-1-2) Generation part The generation unit 192 generates a prediction model M1 that predicts the air conditioning load in the target space TS in the building BL during a third period. In this embodiment, the third period is one week from Sunday to Saturday. The generation unit 192 generates the prediction model M1 after 11:00 PM on Saturday.

[0054] In this embodiment, the prediction model M1 is expressed by the following Equation 1 and Table 1.

[0055]

number

[0056] [Table 1] In Equation 1, the superscript represents time. Therefore, Equation 1 is a prediction model M1 that predicts the air conditioning load at time t+1. Each term on the left side of Equation 1 is further expressed by the following Equations 2 to 8 and Tables 2 to 6.

[0057]

number

[0058]

number

[0059]

number

[0060]

number

[0061]

number

[0062]

number

[0063]

number

[0064] [Table 2] [Table 3] [Table 4] [Table 5] [Table 6] In this embodiment, time t is on the hour, and the time interval between time t and time t+1 is one hour. Therefore, "Δt [s]" shown in equation 6 is 3600 seconds. For example, if the control unit 49 acquires driving data D1 at 0:00, 0:30, 1:00, etc. every 30 minutes, time t may be 0:00, 0:30, 1:00, etc. In this case, the time interval between time t and time t+1 is 30 minutes, and "Δt [s]" is 1800 seconds.

[0065] The "k" in equation 2 indicates the direction of the wall.

[0066] The prediction model M1 particularly has a first parameter, a second parameter, and a third parameter.

[0067] The items shown in Table 2 are defined as second parameters. The second parameters are mainly parameters related to the thermal characteristics of the building BL. The initial values ​​of the second parameters are set by the generation unit 192 and then corrected by the second correction unit 194.

[0068] The items shown in Table 3 are defined as third parameters. The third parameters are mainly parameters related to the heat flow response (to thermal excitation) of the walls of the building BL. In this embodiment, the value of the third parameter is determined using three patterns (multiple patterns) of values ​​that the third parameter can take. The three patterns are determined in advance depending on the magnitude of the heat capacity of the walls of the building BL. FIG. 4 shows, as an example, the TF ow,j 4 shows three patterns of values ​​that TF (j=0 to 20) can take. ow,j (j=0~20) takes three values ​​(straight line 81: small heat capacity, dashed line 82: medium heat capacity, dashed line 83: large heat capacity) depending on the heat capacity of the wall of the building BL. ow,j (j=0~20), AF ow,j (j=0~20), and AF iw,j Since each of j (j=0 to 20) takes three different values, the third parameter takes a total of 27 (=3×3×3) different values. The value of the third parameter is set to an initial value by the generation unit 192 and then corrected by the second correction unit 194.

[0069] The internal heat generation coefficient shown in Table 4 is defined as a first parameter. The internal heat generation coefficient is a parameter that indicates the magnitude of internal heat generation in the target space TS. The internal heat generation coefficient takes a real value between 0 and 1. The internal heat generation coefficient has a value for each of a plurality of times. In this embodiment, the internal heat generation coefficient has a value for each hour on the hour. The value of the internal heat generation coefficient is set to an initial value by the generation unit 192, and is then corrected by the first correction unit 193.

[0070] The values ​​of the items shown in Table 5 are constants. The values ​​of the items shown in Table 5 are set by the generation unit 192. In particular, the mechanical ventilation coefficient has a value for each of a plurality of times. In this embodiment, the mechanical ventilation coefficient has a value for each hour on the hour.

[0071] When the generation unit 192 generates the prediction model M1, or when the first correction unit 193 and the second correction unit 194 correct the prediction model M1, the outdoor temperature, indoor temperature, equivalent outdoor air temperature, and solar radiation of the operating data D1 are input into the outdoor temperature, indoor temperature, equivalent outdoor air temperature, and solar radiation of the items shown in Table 6. When the control device 40 predicts the air conditioning load in the target space TS using the prediction model M1, the outdoor temperature, indoor temperature, equivalent outdoor air temperature, and solar radiation of the time to be predicted are input into the items shown in Table 6.

[0072] First, the generation unit 192 sets the values ​​of the items shown in Table 5 in the prediction model M1. For example, the administrator stores the values ​​of the items shown in Table 5 in advance in the storage unit 11 based on the environment of the target space TS. The generation unit 192 sets the values ​​of the items shown in Table 5, which the administrator has stored in advance in the storage unit 11, in the prediction model M1.

[0073] Next, the generation unit 192 sets initial values ​​of the second parameter and the third parameter in the prediction model M1. The generation unit 192 transforms Equation 1 into the following Equations 9 to 15.

[0074]

number

[0075]

number

[0076]

number

[0077]

number

[0078]

number

[0079]

number

[0080]

number

[0081] The formula 9 is a multiple regression equation with the second parameter as a coefficient. The generation unit 192 generates a learning data set ((x 1 solar ,x 1 skin_win ,x 1 skin_wall ,x 1 storage ,x 1 vent ,SD 1 ig ,Y 1 ),…,(x 128 solar ,x 128 skin_win ,x 128 skin_wall ,x 128 storage ,x 128 vent ,SD 128 ig ,Y 128 )) is prepared. At this time, the generation unit 192 sets one of the 27 values ​​described above to the third parameter. Furthermore, the generation unit 192 inputs the outdoor temperature, indoor temperature, equivalent outdoor temperature, and amount of solar radiation of the operation data D1 at the corresponding time to the outdoor temperature, indoor temperature, equivalent outdoor temperature, and amount of solar radiation of the items shown in Table 6. Furthermore, the generation unit 192 inputs the outdoor temperature, indoor intake temperature, equivalent outdoor temperature, and amount of solar radiation of the operation data D1 at the corresponding time to the items shown in Table 6. Furthermore, the generation unit 192 inputs the air conditioning load Φ HVACThe generation unit 192 inputs the air conditioning load of the operating data D1 for the corresponding time into the target space TS. In addition, the generation unit 192 sets an initial value for the internal heat generation coefficient for the corresponding time. For example, the administrator stores in advance in the storage unit 11 the initial value for the internal heat generation coefficient for each hour on the hour based on the environment of the target space TS and empirical rules. The generation unit 192 sets the initial value for the internal heat generation coefficient for each hour that the administrator stored in advance in the storage unit 11 as the internal heat generation coefficient of the prediction model M1. After preparing the learning dataset, the generation unit 192 calculates the value of the second parameter by the least squares method. In other words, the generation unit 192 calculates the value of the second parameter that minimizes the sum of squares of the difference between the actual measured value of Y and the predicted value of Y.

[0082] The generation unit 192 sequentially sets all 27 values ​​for the third parameter, and calculates the value of the second parameter that minimizes the sum of squares of the differences between the actual measured value of Y and the predicted value of Y 27 times. The generation unit 192 determines the values ​​of the second parameter and the third parameter that minimize the "minimum sum of squares of the differences between the actual measured value of Y and the predicted value of Y" among the 27 "minimum values ​​of the sum of squares of the differences between the actual measured value of Y and the predicted value of Y." The generation unit 192 sets the determined values ​​of the second parameter and the third parameter as initial values ​​in the prediction model M1.

[0083] The generation unit 192 stores the generated prediction model M1 in the storage unit 11.

[0084] (2-2-1-3) First correction unit After the generation unit 192 generates the prediction model M1, the first correction unit 193 corrects the values ​​of the second parameter and the third parameter in the first period, and the second correction unit 194 corrects the value of the internal heat generation coefficient in the second period. In this embodiment, the first period and the second period are one week from Sunday to Saturday.

[0085] During a first period, the first correction unit 193 corrects the values ​​of the second parameter and the third parameter in the prediction model M1 in which the value of the internal heat generation coefficient is set, using the actual measured value of the air conditioning load in the target space TS during the first period. In this embodiment, the first correction unit 193 corrects the values ​​of the second parameter and the third parameter after 11:00 PM on Saturday.

[0086] Similar to the generation unit 192, the first correction unit 193 prepares a learning data set (including actual measured values ​​of the air conditioning load in the target space TS for a first period) for 168 sets of times on the hour from midnight on Sunday to 11:00 pm on Saturday, and determines the values ​​of the second and third parameters. The first correction unit 193 corrects the pre-correction values ​​of the second and third parameters to the determined values ​​of the second and third parameters.

[0087] The first correction unit 193 corrects the values ​​of the second parameter and the third parameter in the prediction model M1 in which the value of the internal heat generation coefficient corrected by the second correction unit 194 in the second period immediately preceding the first period is set. If the second correction unit 194 has not performed correction before the correction by the first correction unit 193, the first correction unit 193 corrects the values ​​of the second parameter and the third parameter in the prediction model M1 in which the initial value of the internal heat generation coefficient is set by the generation unit 192.

[0088] (2-2-1-4) Second correction unit During the second period, the second correction unit 194 corrects the value of the internal heat generation coefficient set in the prediction model M1 using the actual measured value of the air conditioning load in the target space TS for the second period and the predicted value of the air conditioning load for the second period predicted using the prediction model M1 in which the values ​​of the second parameter and the third parameter corrected by the first correction unit 193 are set. In this embodiment, the second correction unit 194 corrects the value of the internal heat generation coefficient after 11 PM on Saturday.

[0089] First, the second correction unit 194 prepares 168 actual measured values ​​of the air conditioning load at the hour from midnight on Sunday to 11:00 pm on Saturday. For example, the generation unit 192 uses the air conditioning load in the operating data D1 at the corresponding time as the actual measured values ​​of the air conditioning load at the hour.

[0090] Next, the second correction unit 194 calculates predicted values ​​of the air conditioning load at 168 times on the hour from midnight on Sunday to 11:00 p.m. on Saturday, using the prediction model M1 in which the values ​​of the second and third parameters corrected by the first correction unit 193 are set. In other words, the second correction unit 194 calculates seven predicted values ​​of the air conditioning load (from Sunday to Saturday) for each time t (on the hour) in the second period. For example, when calculating the predicted value of the air conditioning load at time t, the second correction unit 194 inputs the outdoor temperature, indoor intake temperature, equivalent outdoor air temperature, and amount of solar radiation of the operating data D1 for the times corresponding to the outdoor temperature, indoor temperature, equivalent outdoor air temperature, and amount of solar radiation of the items shown in Table 6, respectively.

[0091] Next, the second correction unit 194 corrects the value of the first parameter at time t based on the difference between the actual measured value of the air conditioning load at time t and the predicted value of the air conditioning load at time t. Specifically, the second correction unit 194 calculates a representative value of the differences at time t using seven differences obtained by subtracting the actual measured value of the air conditioning load at the corresponding time t from each of the seven predicted values ​​of the air conditioning load at time t. The representative value is, for example, an average or a median. If the representative value is positive, the second correction unit 194 corrects the value of the first parameter at time t to a smaller value (because the predicted value is larger than the actual measured value). If the representative value is negative, the second correction unit 194 corrects the value of the first parameter at time t to a larger value (because the predicted value is smaller than the actual measured value).

[0092] The second correction unit 194 determines the amount of correction to the value of the first parameter at time t according to the representative value. For example, the second correction unit 194 calculates "(-1) x (representative value / (floor area A x internal heat generation i)" as follows: g))" is used as the correction amount. In this case, an upper limit may be set on the absolute value of the correction amount so that the value of the internal heat generation coefficient to which the correction amount has been added is a real number between 0 and 1. Also, for example, the second correction unit 194 may set the correction amount to ±0.1 when the absolute value of the representative value / floor area A is greater than a predetermined value, and may set the correction amount to zero when the absolute value of the representative value / floor area A is smaller than the predetermined value. The predetermined value may be, for example, "internal heat generation i g ×0.1".

[0093] (2-2-1-5) Transmitter The transmission unit 195 transmits the prediction model M1 to the control device 40 every time the first correction unit 193 or the second correction unit 194 corrects the prediction model M1.

[0094] (3) Processing An example of the processing of the tuning device 1 will be described with reference to the flowchart of Fig. 5. As a premise, the tuning device 1 acquires operating data D1 at the top of the hour from the control device 40 every hour.

[0095] As shown in step S1, the tuning device 1 generates a prediction model M1 after 11:00 p.m. on Saturday during a week (third period) from Sunday to Saturday, using 168 sets of learning data sets at the hour from midnight on Sunday to 11:00 p.m. on Saturday.

[0096] After completing step S1, as shown in step S2, the tuning device 1 corrects the values ​​of the second and third parameters of the prediction model M1 after 11:00 p.m. on Saturday in the next week (first period) using 168 sets of learning data sets at the hour from midnight on Sunday to 11:00 p.m. on Saturday.

[0097] After completing step S2, the tuning device 1 corrects the values ​​of the second and third parameters of the prediction model M1 and then transmits the corrected prediction model M1 to the control device 40, as shown in step S3.

[0098] After completing step S3, as shown in step S4, the tuning device 1 corrects the value of the first parameter of the prediction model M1 after 11:00 p.m. on Saturday in the next week (second period) based on the actual measured values ​​and predicted values ​​of the air conditioning load at 168 points on the hour from midnight on Sunday to 11:00 p.m. on Saturday.

[0099] After completing step S4, the tuning device 1 corrects the value of the first parameter of the prediction model M1 and then transmits the corrected prediction model M1 to the control device 40, as shown in step S5.

[0100] Thereafter, the tuning device 1 repeats the following steps every week: correcting the values ​​of the second and third parameters of the prediction model M1 (step S2), transmitting the corrected prediction model M1 to the control device 40 (step S3), correcting the value of the first parameter of the prediction model M1 (step S4), and transmitting the corrected prediction model M1 to the control device 40 (step S5).

[0101] (4) Features (4-1) 2. Description of the Related Art Conventionally, there is a technique for predicting the air conditioning load in a target space using a prediction model in which a parameter value indicating the magnitude of internal heat generation is set.

[0102] The magnitude of internal heat generation changes depending on the season. Therefore, in order to more accurately predict the air conditioning load in the target space using the prediction model, it is desirable to appropriately correct the value of the parameter indicating the magnitude of internal heat generation set in the prediction model.

[0103] The tuning method of this embodiment is performed by a tuning device 1. In the tuning method, a prediction model M1 is tuned. The prediction model M1 predicts the air conditioning load in a target space TS within a building BL. The prediction model M1 has an internal heat generation coefficient (first parameter) and a second parameter. The internal heat generation coefficient (first parameter) is a parameter indicating the magnitude of internal heat generation in the target space TS. The second parameter is a parameter related to the thermal characteristics of the building BL. The tuning method includes a first step and a second step. In the first step, an actual measurement value of the air conditioning load in the target space TS for a first period is used to correct the value of the second parameter in the prediction model M1, in which the value of the internal heat generation coefficient (first parameter) is set. In the second step, an actual measurement value of the air conditioning load in the target space TS for a second period is used to correct the value of the internal heat generation coefficient (first parameter) set in the prediction model M1, in which the value of the second parameter corrected by performing the first step is set.

[0104] In the tuning method of this embodiment, the prediction model M1 has an internal heat generation coefficient (first parameter) and a second parameter. The internal heat generation coefficient (first parameter) is a parameter that indicates the magnitude of internal heat generation in the target space TS. The second parameter is a parameter related to the thermal characteristics of the building BL. The tuning method corrects the value of the internal heat generation coefficient (first parameter) in accordance with the correction of the value of the second parameter. As a result, the prediction model M1 tuned by the tuning method can more accurately predict the air conditioning load in the target space TS.

[0105] (4-2) In the tuning method of this embodiment, the tuning device 1 performs a first step in a first period, and a second step in a second period.

[0106] (4-3) In the tuning method of this embodiment, the tuning device 1 repeatedly performs a first step in a first period and a second step in a second period. The first step corrects the value of a second parameter in the prediction model M1, which has the value of the internal heat generation coefficient (first parameter) corrected by performing the second step in a second period immediately before the first period.

[0107] As a result, the tuning method can efficiently improve the accuracy of the prediction model M1 by repeating the correction of the value of the first parameter and the correction of the value of the second parameter.

[0108] (4-4) In the tuning method of this embodiment, the internal heat generation coefficient (first parameter) has a value for each of a plurality of times including a first time. In a second step, a predicted value of the air conditioning load at the first time is calculated using a prediction model M1. In the second step, the value of the internal heat generation coefficient (first parameter) at the first time is corrected based on the difference between the actual measured value of the air conditioning load at the first time and the predicted value of the air conditioning load at the first time.

[0109] As a result, by the tuning method correcting the value of the internal heat generation coefficient (first parameter) at multiple times, the prediction model M1 tuned by the tuning method can predict the air conditioning load in the target space TS more precisely.

[0110] (4-5) In the tuning method of this embodiment, the second step calculates a plurality of predicted values ​​of air conditioning loads at a first time in a second period. The second step calculates a representative value of the differences at the first time using a plurality of differences obtained by subtracting the corresponding actual measured values ​​of the air conditioning loads at the first time from the calculated predicted values ​​of the air conditioning loads at the first time. If the representative value is positive, the second step corrects the value of the internal heat generation coefficient (first parameter) at the first time to a smaller value. If the representative value is negative, the second step corrects the value of the internal heat generation coefficient (first parameter) at the first time to a larger value.

[0111] As a result, the tuning method can correct the value of the internal heat generation coefficient (first parameter) without being biased towards the magnitude of internal heat generation on a particular day.

[0112] (4-6) In the tuning method of this embodiment, the second step determines the amount of correction for the value of the internal heat generation coefficient (first parameter) at the first time point, depending on the representative value.

[0113] (4-7) In the tuning method of this embodiment, the second period is one week.

[0114] As a result, the tuning method can correct the value of the internal heat generation coefficient (first parameter) without biasing towards the magnitude of internal heat generation on a particular day of the week.

[0115] (4-8) In the tuning method of this embodiment, the prediction model M1 further includes a third parameter. The third parameter is a parameter related to the heat flow response of the wall of the building BL. In the first step, the value of the third parameter is corrected using multiple patterns of values ​​that the third parameter can take. The multiple patterns are determined in advance according to the magnitude of the heat capacity of the wall.

[0116] As a result, the tuning method can efficiently correct the value of the third parameter.

[0117] (4-9) The control device 40 of this embodiment includes a control unit 49. The control unit 49 predicts the air conditioning load in the target space TS using a prediction model M1 tuned by a tuning method. The control unit 49 determines a control target value for the indoor unit 20 installed in the target space TS based on the predicted air conditioning load in the target space TS.

[0118] (4-10) The tuning device 1 of this embodiment tunes a prediction model M1. The prediction model M1 predicts the air conditioning load in a target space TS within a building BL. The tuning device 1 includes a control unit 19. The prediction model M1 has an internal heat generation coefficient (first parameter) and a second parameter. The internal heat generation coefficient (first parameter) is a parameter indicating the magnitude of internal heat generation in the target space TS. The second parameter is a parameter related to the thermal characteristics of the building BL. The control unit 19 corrects the value of the second parameter in the prediction model M1, in which the value of the internal heat generation coefficient (first parameter) is set, using an actual measurement value of the air conditioning load in the target space TS for a first period. The control unit 19 corrects the value of the internal heat generation coefficient (first parameter) set in the prediction model M1, using an actual measurement value of the air conditioning load in the target space TS for a second period and a predicted value of the air conditioning load for the second period predicted using the prediction model M1 in which the corrected value of the second parameter is set.

[0119] In the tuning device 1 of this embodiment, the prediction model M1 has an internal heat generation coefficient (first parameter) and a second parameter. The internal heat generation coefficient (first parameter) is a parameter that indicates the magnitude of internal heat generation in the target space TS. The second parameter is a parameter related to the thermal characteristics of the building BL. The tuning device 1 corrects the value of the internal heat generation coefficient (first parameter) in accordance with the correction of the value of the second parameter. As a result, the prediction model M1 tuned by the tuning device 1 can more accurately predict the air conditioning load in the target space TS.

[0120] (5) Variations (5-1) Variation 1A In this embodiment, one indoor unit 20 is installed in the target space TS. However, multiple indoor units 20 may be installed in the target space TS.

[0121] At this time, the operating data D1 includes the air volume of the indoor fan, the opening degree of the indoor expansion valve 23, the indoor suction temperature, the indoor heat exchanger temperature, and the air conditioning load of each indoor unit 20.

[0122] For example, when a prediction model M1 is generated by the generation unit 192, or when the prediction model M1 is corrected by the first correction unit 193 and the second correction unit 194, the indoor temperature and air conditioning load of the prediction model M1 are input with the average indoor intake temperature and the total air conditioning load of the operating data D1, respectively.

[0123] For example, the control unit 49 uses a prediction model M1 to predict the air conditioning load of each indoor unit 20. The control unit 49 inputs the outdoor temperature, indoor temperature, equivalent outdoor temperature, and solar radiation at the time of the desired prediction, as well as past indoor temperatures, into the prediction model M1, and predicts the air conditioning load in the target space TS. The control unit 49 allocates the predicted air conditioning load in the target space TS proportionally to each indoor unit 20 based on the position of each indoor unit 20, etc.

[0124] (5-2) Although the embodiments of the present disclosure have been described above, it will be understood that various changes in form and details can be made without departing from the spirit and scope of the present disclosure as defined in the claims. [Explanation of symbols]

[0125] 1 Tuning device 19 Control Unit 20 Indoor unit 40 Control device 49 Control Unit BL Building M1 predictive model TS target space [Prior art documents] [Patent documents]

[0126] [Patent Document 1] Patent No. 7125644

Claims

1. A tuning method for tuning a prediction model (M1) that predicts an air conditioning load in a target space (TS) in a building (BL), performed by a tuning device (1), comprising: The prediction model has a first parameter indicating the magnitude of internal heat generation in the target space and a second parameter related to the thermal characteristics of the building, a first step of correcting the value of the second parameter in the prediction model in which the value of the first parameter is set, using an actual measurement value of an air conditioning load in the target space for a first period; a second step of correcting the value of the first parameter set in the prediction model using an actual measurement value of the air conditioning load for a second period in the target space and a predicted value of the air conditioning load for the second period predicted using the prediction model in which the value of the second parameter corrected by performing the first step is set; Equipped with Tuning method.

2. the tuning device repeats performing the first step in the first period and performing the second step in the second period; the first step corrects the value of the second parameter in the prediction model in which the value of the first parameter corrected by performing the second step in the second period immediately before the first period is set; The tuning method of claim 1 .

3. the first parameter has a value for each of a plurality of times including a first time; The second step is calculating a predicted value of the air conditioning load at the first time using the prediction model; correcting the value of the first parameter at the first time based on a difference between an actual measured value of the air conditioning load at the first time and a predicted value of the air conditioning load at the first time; 3. The tuning method according to claim 1 or 2.

4. The second step is calculating predicted values ​​of a plurality of air conditioning loads at the first time in the second period; calculating a representative value of the differences at the first time using a plurality of differences obtained by subtracting the corresponding actual measured values ​​of the air conditioning loads at the first time from the calculated predicted values ​​of the air conditioning loads at the first time; If the representative value is positive, correcting the value of the first parameter at the first time to a smaller value; If the representative value is negative, the value of the first parameter at the first time is corrected to a larger value. The tuning method according to claim 3 .

5. the second step determines a correction amount for the value of the first parameter at the first time in accordance with the representative value; 5. The tuning method according to claim 4.

6. The second period is one week.

5. The tuning method according to claim 4.

7. the predictive model further comprises a third parameter related to the heat flow response of the wall of the building; the first step corrects the value of the third parameter using a plurality of patterns of values ​​that the third parameter can take, which patterns are determined in advance according to the magnitude of the heat capacity of the wall; 3. The tuning method according to claim 1 or 2.

8. A control unit (49) is provided, The control unit predicting an air conditioning load in the target space using the prediction model tuned by the tuning method according to claim 1 or 2; determining a control target value for an indoor unit (20) installed in the target space based on the predicted air conditioning load in the target space; A control device (40).

9. A tuning device (1) that tunes a prediction model (M1) that predicts an air conditioning load in a target space (TS) in a building (BL), A control unit (19) is provided, The prediction model has a first parameter indicating the magnitude of internal heat generation in the target space and a second parameter related to the thermal characteristics of the building, The control unit correcting the value of the second parameter in the prediction model in which the value of the first parameter is set using an actual measurement value of the air conditioning load in the target space for the first period; correcting the value of the first parameter set in the prediction model using an actual measurement value of the air conditioning load for the second period in the target space and a predicted value of the air conditioning load for the second period predicted using the prediction model in which the corrected value of the second parameter is set; Tuning device (1).

Images