Control system
The control system addresses inefficiencies in facility equipment control by generating dummy operating data and simulating control models, ensuring efficient operation despite site-specific variations.
Patent Information
- Application Number
- JP2024021694
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-02-16
- Publication Date
- 2025-08-28
AI Technical Summary
Existing control systems for facility equipment struggle to efficiently learn and apply control models when the equipment configuration varies significantly by installation site, leading to inefficiencies in operation.
A control system that generates dummy operating data based on on-site information and uses a theoretical physical model to simulate and learn a control model before equipment operation, allowing for efficient control even with varying configurations.
Enables efficient control of facility equipment from the start of operation by generating and learning control models using dummy data, ensuring accurate and efficient operation despite site-specific variations.
Smart Images

Figure 2025125642000001_ABST
Abstract
Description
[Technical Field]
[0001] Regarding control systems. [Background technology]
[0002] As shown in Patent Document 1 (JP 2022-145655 A), there is a technology that, when there is no operating data for an equipment device, uses operating data for equipment similar to the equipment device to learn a control model for controlling the equipment device. Summary of the Invention [Problem to be solved by the invention]
[0003] When the configuration of facility equipment varies greatly depending on the site where it is installed, Patent Document 1 does not allow the use of operating data of equipment similar to the facility equipment, and therefore does not allow the control model to be learned before the facility equipment starts to operate. Therefore, when the configuration of facility equipment varies greatly depending on the site where it is installed, Patent Document 1 has a problem in that it is not possible to efficiently control the facility equipment using the control model when the facility equipment starts to operate. [Means for solving the problem]
[0004] A control system according to a first aspect includes an equipment and a control unit. The control unit controls the equipment. The control unit generates first operating data based on on-site information. The on-site information is information about the equipment at a site where the equipment is installed. The first operating data is dummy operating data for the equipment. The control unit uses the first operating data to learn a first control model. The control unit controls the equipment using the first control model.
[0005] In the control system of the first aspect, the control unit generates first operating data based on local information. The local information is information about the equipment at the local site where the equipment is installed. The first operating data is dummy operating data for the equipment. Therefore, even if the configuration of the equipment varies significantly depending on the local site where it is installed, the control system can generate the first operating data for the equipment and learn the first control model before the equipment starts to operate. As a result, the control system can efficiently control the equipment using the first control model from the start of operation of the equipment, even if the configuration of the equipment varies significantly depending on the local site where it is installed.
[0006] A control system according to a second aspect is the control system according to the first aspect, wherein the on-site information includes drawing information of the facility equipment and characteristic information of the facility equipment, and the control unit generates first operating data based on the on-site information by a simulation using a theoretical physical model.
[0007] With this configuration, the control system according to the second aspect can generate a large amount of various types of operating data.
[0008] A control system according to a third aspect is the control system according to the second aspect, wherein the control unit acquires second operating data. The second operating data is actual operating data of the facility equipment. The control unit corrects on-site information using the second operating data. The control unit generates third operating data by simulation based on the corrected on-site information. The third operating data is operating data of the facility equipment. The control unit learns a second control model using the third operating data. The control unit controls the facility equipment using the second control model.
[0009] In the control system of the third aspect, third operating data is generated by simulation based on the corrected on-site information. The control unit uses the third operating data to learn a second control model. The control unit controls the facility equipment using the second control model. As a result, the control system can control the facility equipment with higher accuracy.
[0010] A control system according to a fourth aspect is the control system according to any one of the first aspect to the third aspect, wherein the facility equipment includes a first pipe and one or more first devices. The first pipe transports a medium. The first devices control the flow of the medium. The drawing information includes an arrangement of the first pipe and an arrangement of the first devices. The arrangement of the first pipe includes the length, height, and shape of the first pipe. The control unit controls the first device using the first control model or the second control model.
[0011] With this configuration, the control system according to the fourth aspect can automatically control one or more first appliances, for example, to improve energy conservation and comfort.
[0012] A fifth aspect of the control system is the control system of the fourth aspect, wherein the facility equipment is an air conveying system. The first pipe is a duct. The medium is air. The first equipment includes a fan. The operation data of the facility equipment includes the fan air volume, the fan static pressure, and the fan frequency. The control unit controls the fan air volume using the first control model or the second control model.
[0013] With such a configuration, the control system according to the fifth aspect can automatically control the balance of airflow rates of a plurality of fans so as to improve energy conservation and comfort, for example.
[0014] A sixth aspect of the control system is the fourth aspect of the control system, wherein the facility equipment is a water conveyance system. The first pipe is a water pipe. The medium is water. The first equipment includes a pump. The operation data of the facility equipment includes a pump flow rate, a pump pressure, and a pump frequency. The control unit controls the pump flow rate using the first control model or the second control model.
[0015] With such a configuration, the control system according to the sixth aspect can automatically control the balance of flow rates of a plurality of pumps so as to improve energy conservation and comfort, for example. [Brief explanation of the drawings]
[0016] [Figure 1] FIG. 1 is a schematic diagram of a water transport system. [Figure 2] FIG. 1 is a schematic diagram of an air conveying system. [Figure 3] FIG. 2 is a control block diagram of the water transport system. [Figure 4] FIG. 2 is a control block diagram of the air conveying system. [Figure 5] FIG. 2 is a control block diagram of a control device. [Figure 6] FIG. 1 is a diagram showing an example of a pressure-flow rate characteristic curve. [Figure 7] FIG. 1 is a diagram showing an example of a static pressure vs. air volume characteristic curve. [Figure 8] 1 is a flowchart illustrating an example of processing in a water delivery system. [Figure 9] 1 is a flowchart illustrating an example of processing in an air distribution system. DETAILED DESCRIPTION OF THE INVENTION
[0017] (1) Overall structure The control system 1 controls the flow of a medium in a medium conveyance system (facility equipment). In this embodiment, the control system 1 controls the flow of water in a water conveyance system 2 that conveys water as a medium. The control system 1 also controls the flow of air in an air conveyance system 3 that conveys air as a medium. FIG. 1 is a schematic configuration diagram of the water conveyance system 2. FIG. 2 is a schematic configuration diagram of the air conveyance system 3. As shown in FIGS. 1 and 2, the control system 1 includes the water conveyance system 2, the air conveyance system 3, and a control device 9. The water conveyance system 2 cools or heats water flowing in the first water circuit WC1 using refrigerant flowing through refrigerant circuits RCa and RCb. The air conveyance system 3 air-conditions two air-conditioning zones Z1 and Z2 in a room RM using the water flowing through the first water circuit WC1 that has been cooled or heated by the water conveyance system 2. The number of rooms and air-conditioning zones air-conditioned by the air conveyance system 3 can be designed to be one or more, as appropriate.
[0018] (2) Detailed configuration (2-1) Water transport system As shown in Figure 1, the water transport system 2 mainly includes chiller units 10a, 10b, a first water circuit WC1, a second water circuit WC2a, WC2b, inlet temperature sensors 81a, 81b, outlet temperature sensors 82a, 82b, a supply water temperature sensor 83, a return water temperature sensor 84, and a heat source side control device 19.
[0019] (2-1-1) Chiller unit 1, chiller units 10a and 10b cool or heat the water flowing in the first water circuit WC1. The number of chiller units included in the water transport system 2 may be one or more, as appropriate.
[0020] The chiller units 10a and 10b mainly include refrigerant circuits RCa and RCb, and first pumps 15a and 15b (first devices).
[0021] The refrigerant circuits RCa and RCb mainly include compressors 11a and 11b, flow path switching mechanisms 18a and 18b, second heat exchangers 12a and 12b, expansion valves 13a and 13b, and first heat exchangers 14a and 14b.
[0022] The compressors 11a and 11b draw in low-pressure refrigerant, compress the refrigerant using a compression mechanism (not shown), and discharge the compressed refrigerant. The compressors 11a and 11b are, for example, rotary or scroll type positive displacement compressors. The frequencies of the compressors 11a and 11b can be controlled by inverters.
[0023] The flow path switching mechanisms 18a, 18b switch the flow path of the refrigerant depending on whether the water flowing in the heat source side first water circuits WC11a, WC11b is to be cooled or heated.
[0024] The second heat exchangers 12a and 12b exchange heat between the refrigerant flowing in the refrigerant circuits RCa and RCb and the refrigerant flowing in the second water circuits WC2a and WC2b. The second heat exchangers 12a and 12b are, for example, fin-and-tube heat exchangers having a plurality of heat transfer fins and a plurality of heat transfer tubes.
[0025] The expansion valves 13a and 13b are mechanisms for adjusting the pressure and flow rate of the refrigerant flowing in the refrigerant circuits RCa and RCb, and the opening degrees of the expansion valves 13a and 13b are controllable.
[0026] The first heat exchangers 14a, 14b exchange heat between the refrigerant flowing in the refrigerant circuits RCa, RCb and the water flowing in the heat-source-side first water circuits WC11a, WC11b. The first heat exchangers 14a, 14b are, for example, fin-and-tube heat exchangers having a plurality of heat transfer fins and a plurality of heat transfer tubes.
[0027] The first pumps 15a and 15b are installed in the water pipes WP3a and WP3b. The first pumps 15a and 15b circulate the water in the first water circuit WC1 by delivering the water in the water pipes WP3a and WP3b in the directions indicated by the arrows in FIG. 1. In other words, the first pumps 15a and 15b control the flow of water. The frequency of the first pumps 15a and 15b can be controlled by an inverter.
[0028] (2-1-2)Second water circuit As shown in FIG. 1, the second water circuits WC2a and WC2b include second heat exchangers 12a and 12b, third pumps 16a and 16b, and cooling towers 17a and 17b.
[0029] The third pumps 16a, 16b circulate the water in the second water circuits WC2a, WC2b in the directions shown by the arrows in Figure 1. The frequencies of the third pumps 16a, 16b can be controlled by an inverter.
[0030] The cooling towers 17a and 17b cool the water flowing in the second water circuits WC2a and WC2b.
[0031] (2-1-3) 1st water circuit 1, the first water circuit WC1 mainly includes a supply header 21, a second pump 22 (first device), a third heat exchanger 34 of the air handling unit 30, a flow control valve 23, a return header 24, first pumps 15a and 15b, and first heat exchangers 14a and 14b. The number of pumps included in the water transport system 2 may be one or more, as appropriate.
[0032] The supply header 21 and the third heat exchanger 34 are connected by a water pipe WP1 (first pipe). The third heat exchanger 34 and the return header 24 are connected by a water pipe WP2 (first pipe). The return header 24 and the first heat exchangers 14a, 14b are connected by water pipes WP3a, WP3b (first pipe). The first heat exchangers 14a, 14b and the supply header 21 are connected by water pipes WP4a, WP4b (first pipe). The supply header 21 and the return header 24 are connected by a water pipe WP5 (first pipe). The water pipes WP1 to WP5 transport water.
[0033] The water pipe WP1, the third heat exchanger 34, and the water pipe WP2 constitute the user-side first water circuit WC12. The water pipes WP3a and WP3b, the first heat exchangers 14a and 14b, and the water pipes WP4a and WP4b constitute the heat-source-side first water circuits WC11a and WC11b. The water pipe WP5 constitutes a bypass circuit. The bypass circuit is installed to return water sent in excess to the forward header section 21 to the return header section 24.
[0034] The forward header section 21 straightens the water that has passed through the heat source side first water circuits WC11a and WC11b, and sends it out to the utilization side first water circuit WC12.
[0035] The second pump 22 is installed in the water pipe WP1. The second pump 22 circulates the water in the first water circuit WC1 by delivering the water in the water pipe WP1 in the direction shown by the arrow in Fig. 1. In other words, the second pump 22 controls the flow of water. The frequency of the second pump 22 can be controlled by an inverter.
[0036] The flow rate adjustment valve 23 is installed in the water pipe WP2. The flow rate adjustment valve 23 is a mechanism for adjusting the pressure and flow rate of the water flowing through the water pipe WP2. The opening degree of the flow rate adjustment valve 23 is controllable.
[0037] The return header section 24 straightens the water that has passed through the utilization side first water circuit WC12 and sends it out to the heat source side first water circuits WC11a and WC11b.
[0038] (2-1-4) Sensor 1, inlet temperature sensors 81a and 81b are installed in water pipes WP3a and WP3b, and measure the temperature of water flowing through water pipes WP3a and WP3b before flowing into chiller units 10a and 10b.
[0039] The outlet temperature sensors 82a and 82b are installed on the water pipes WP4a and WP4b, and measure the temperatures of the water flowing out of the chiller units 10a and 10b and flowing through the water pipes WP4a and WP4b.
[0040] The forward water temperature sensor 83 is installed in the forward header section 21. The forward water temperature sensor 83 measures the temperature of the water flowing out from the forward header section 21 into the utilization side first water circuit WC12.
[0041] The return water temperature sensor 84 is installed in the return header section 24. The return water temperature sensor 84 measures the temperature of the water flowing into the return header section 24 from the utilization side first water circuit WC12.
[0042] (2-1-5) Heat source side control device The heat source side control device 19 controls the operation of each component of the water transport system 2. Figure 3 is a control block diagram of the water transport system 2. As shown in Figure 3, the heat source side control device 19 is communicatively connected to the compressors 11a, 11b, flow path switching mechanisms 18a, 18b, expansion valves 13a, 13b, first pumps 15a, 15b, second pump 22, third pumps 16a, 16b, flow control valve 23, inlet temperature sensors 81a, 81b, outlet temperature sensors 82a, 82b, supply water temperature sensor 83, and return water temperature sensor 84 so as to exchange control signals and the like.
[0043] The heat source side control device 19 has a control and arithmetic device and a storage device. The control and arithmetic device is a processor such as a CPU or GPU. The storage device is a storage medium such as a RAM, a ROM, or a flash memory. The control and arithmetic device reads out a program stored in the storage device and performs predetermined arithmetic processing in accordance with the program, thereby controlling the operation of each component constituting the water transport system 2. The control and arithmetic device can also write calculation results to the storage device and read out information stored in the storage device in accordance with the program.
[0044] The heat source side control device 19 is communicably connected to the use side control device 39 of the air conveyance system 3 and the control device 9 via the network NW so as to exchange control signals and the like.
[0045] (2-2) Pneumatic conveying system 2, the air conveying system 3 mainly includes a use-side unit 30, an outside air duct 41, a return air duct 42, a supply air duct 43 (first pipe), dispersion fans 441, 442 (first equipment), a supply air temperature sensor 71, a supply air humidity sensor 72, a return air temperature sensor 73, a return air humidity sensor 74, indoor temperature sensors 751, 752, indoor humidity sensors 761, 762, air volume sensors 771, 772, a discharge pressure sensor 78, an outside air temperature sensor 61, an outside air humidity sensor 62, and a use-side control device 39. The outside air duct 41, the return air duct 42, and the supply air duct 43 convey air.
[0046] (2-2-1) User unit As shown in FIG. 2, the user-side unit 30 conditions the air in the room RM by using water that has been cooled or heated by the water conveying system 2 and flows through the first water circuit WC1. In this embodiment, the user-side unit 30 is an air handling unit 30. However, the user-side unit 30 may also be a fan coil unit or the like. The number of user-side units 30 included in the air conveying system 3 may be one or more as appropriate. When there are multiple user-side units 30, the multiple user-side units 30 are installed in parallel in the user-side first water circuit WC12.
[0047] Air handling unit 30 has a casing 31. Inside casing 31, there are mainly installed an outside air fan 32, a filter 33, a third heat exchanger 34, a sprinkler humidifier 35, and an air supply fan 36 (first device). The number of fans included in air conveying system 3 is designed to be one or more as appropriate.
[0048] The outdoor air fan 32 draws in air from the outdoor air duct 41 and blows it out into the casing 31. The supply air fan 36 draws in air from the casing 31 and blows it out into the supply air duct 43. This generates an air flow from the outdoor air duct 41 and return air duct 42 side toward the supply air duct 43 side inside the casing 31. In other words, the outdoor air fan 32 and the supply air fan 36 control the air flow. The frequencies of the outdoor air fan 32 and the supply air fan 36 can be controlled by an inverter.
[0049] The filter 33 removes dust particles and other particles contained in the air flow generated within the casing 31 .
[0050] The third heat exchanger 34 exchanges heat between the water flowing in the user-side first water circuit WC12 and the air flowing inside the casing 31. The third heat exchanger 34 is, for example, a fin-and-tube heat exchanger having a plurality of heat transfer fins and a plurality of heat transfer tubes.
[0051] The sprinkler humidifier 35 humidifies the air flowing inside the casing 31 by sprinkling water from a tank (not shown) installed outside the casing 31 into the casing 31 from a nozzle.
[0052] (2-2-2) Duct 2, one end of the outside air duct 41 is connected to the casing 31. The other end of the outside air duct 41 is connected to the outdoors. Outdoor air flows into the casing 31 through the outside air duct 41.
[0053] One end of the return air duct 42 is connected to the casing 31. The other end of the return air duct 42 is connected to the room RM. Air from the room RM flows into the casing 31 through the return air duct 42.
[0054] The air supply duct 43 has a main duct 433 and branch ducts 431 and 432. One end of the main duct 433 is connected to the casing 31. The main duct 433 is also connected to dispersion fans 441 and 442. One ends of the branch ducts 431 and 432 are connected to the dispersion fans 441 and 442. The other ends of the branch ducts 431 and 432 are connected to air-conditioning zones Z1 and Z2 in the room RM. Air conditioned inside the casing 31 flows into the air-conditioning zones Z1 and Z2 in the room RM through the main duct 433 and the branch ducts 431 and 432.
[0055] (2-2-3) Dispersion fan 2, the dispersion fans 441, 442 draw in air from the main duct 433 and blow it out to the air-conditioning zones Z1, Z2 in the room RM through the branch ducts 431, 432. In other words, the dispersion fans 441, 442 control the air flow. The frequency of the dispersion fans 441, 442 can be controlled by an inverter.
[0056] (2-2-4) Sensor 2, supply air temperature sensor 71 and supply air humidity sensor 72 are installed near supply air duct 43 inside casing 31. Supply air temperature sensor 71 measures the temperature of air blown out from inside casing 31 into supply air duct 43. Supply air humidity sensor 72 measures the humidity of air blown out from inside casing 31 into supply air duct 43.
[0057] The return air temperature sensor 73 and the return air humidity sensor 74 are installed near the return air duct 42 inside the casing 31. The return air temperature sensor 73 measures the temperature of the air flowing into the casing 31 from the return air duct 42. The return air humidity sensor 74 measures the humidity of the air flowing into the casing 31 from the return air duct 42.
[0058] Indoor temperature sensors 751 and 752 are installed in air conditioning zones Z1 and Z2 and measure the temperatures of the air in air conditioning zones Z1 and Z2.
[0059] Indoor humidity sensors 761 and 762 are installed in air conditioning zones Z1 and Z2. The indoor humidity sensors 761 and 762 measure the humidity of the air in air conditioning zones Z1 and Z2.
[0060] The air flow sensors 771 and 772 are installed on the dispersion fans 441 and 442. The air flow sensors 771 and 772 measure the air flow rates of the dispersion fans 441 and 442.
[0061] Discharge pressure sensor 78 is installed near the outlet of air supply fan 36. Discharge pressure sensor 78 measures the blowing pressure (static pressure) of air supply fan 36.
[0062] The outdoor air temperature sensor 61 is installed outdoors and measures the temperature of the outdoor air.
[0063] The outdoor humidity sensor 62 is installed outdoors and measures the humidity of the outdoor air.
[0064] (2-2-5) User-side control device The use-side control device 39 controls the operation of each component of the air conveying system 3. Fig. 4 is a control block diagram of the air conveying system 3. As shown in Fig. 4, the use-side control device 39 is communicatively connected to the outdoor air fan 32, the supply air fan 36, the filter 33, the sprinkler humidifier 35, the dispersion fans 441, 442, the supply air temperature sensor 71, the supply air humidity sensor 72, the return air temperature sensor 73, the return air humidity sensor 74, the indoor temperature sensors 751, 752, the indoor humidity sensors 761, 762, the air volume sensors 771, 772, the discharge pressure sensor 78, the outdoor air temperature sensor 61, and the outdoor air humidity sensor 62 so as to exchange control signals and the like.
[0065] The user-side control device 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 out a program stored in the storage device and performs predetermined arithmetic processing in accordance with the program, thereby controlling the operation of each component constituting the air conveying system 3. The control and arithmetic device can also write the results of calculations to the storage device and read out information stored in the storage device in accordance with the program.
[0066] The user-side control device 39 is configured to be able to receive various signals transmitted from the operation remote controller. The various signals include, for example, signals instructing the start and stop of operation and signals related to various settings. The signals related to various settings include, for example, signals related to the set temperature and set humidity.
[0067] The use-side control device 39 is communicably connected to the heat-source-side control device 19 of the water transport system 2 and the control device 9 via the network NW so as to exchange control signals and the like.
[0068] (2-3) Control device 999 is communicatively connected via a network NW to the heat source side control device 19 and the use side control device 39 to exchange control signals and the like. The control device 9 controls the water transport system 2 and the air transport system 3 via the heat source side control device 19 and the use side control device 39.
[0069] 5 is a control block diagram of the control device 9. As shown in FIG. 5, the control device 9 mainly includes a storage unit 91, an input unit 92, a display unit 93, a communication unit 94, and a control unit 99.
[0070] (2-3-1) Storage section The storage unit 91 is a storage device such as a RAM, a ROM, and a HDD, etc. The storage unit 91 stores programs executed by the control unit 99, data necessary for executing the programs, and the like.
[0071] The local information 52 and 53 is stored in advance in the storage unit 91. The local information 52 and 53 is stored using the input unit 92, for example.
[0072] The site information 52 is information about the water transportation system 2 at the site where the water transportation system 2 is installed. The site information 52 includes drawing information 521 of the water transportation system 2 and characteristic information 522 of the water transportation system 2.
[0073] The site information 53 is information about the air conveying system 3 at the site where the air conveying system 3 is installed. The site information 53 includes drawing information 531 of the air conveying system 3 and characteristic information 532 of the air conveying system 3.
[0074] For example, the drawing information 521 of the water transport system 2 includes the arrangement of the water pipes WP1 to WP5 and the arrangement of the first pumps 15a, 15b, and the second pump 22. The arrangement of the water pipes WP1 to WP5 includes the length, height, and shape of the water pipes WP1 to WP5. The height of the water pipes WP1 to WP5 is, for example, the elevation difference between the water pipes WP1 to WP5. The shape of the water pipes WP1 to WP5 is, for example, the bends of the water pipes WP1 to WP5, the diameter of the water pipes WP1 to WP5, etc.
[0075] For example, the drawing information 531 of the air conveying system 3 includes the layout of the air supply duct 43, and the layout of the air supply fan 36 and the distribution fans 441 and 442. The layout of the air supply duct 43 includes the length, height, and shape of the air supply duct 43 (the length, height, and shape of the main duct 433 and the branch ducts 431 and 432). The height of the air supply duct 43 is, for example, the elevation difference of the air supply duct 43. The shape of the air supply duct 43 is, for example, the bends of the air supply duct 43, the diameter of the air supply duct 43, the type of the air supply duct 43 (round, rectangular, or flexible), and the shape of the air outlet of the air supply duct 43 in the room RM.
[0076] For example, characteristic information 522 of water conveyance system 2 is a pressure-flow rate characteristic curve diagram for each of first pumps 15a, 15b and second pump 22. FIG. 6 is a diagram showing an example of a pump pressure-flow rate characteristic curve diagram. As shown in FIG. 6, in the pressure-flow rate characteristic curve diagram, the vertical axis represents pump pressure (P) and the horizontal axis represents pump flow rate (Q). Curves C1 and C2 are pump pressure-flow rate characteristic curves. Curve C1 is the pump pressure-flow rate characteristic curve when the pump frequency is 40 Hz. Curve C2 is the pump pressure-flow rate characteristic curve when the pump frequency is 50 Hz. Curves C3 and C3' are water piping resistance curves. Coefficients A and A' of curves C3 and C3' are determined according to the predetermined length, height, and shape of the water piping extending downstream of the pump.
[0077] For example, characteristic information 532 of air conveying system 3 is a static pressure vs. air volume characteristic curve diagram for each of supply air fan 36 and dispersion fans 441, 442. FIG. 7 is a diagram showing an example of a static pressure vs. air volume characteristic curve diagram for a fan. As shown in FIG. 7, in the static pressure vs. air volume characteristic curve diagram, the vertical axis represents the static pressure (P) of the fan and the horizontal axis represents the air volume (Q) of the fan. Curves C4 and C5 are the static pressure vs. air volume characteristic curves for the fan. Curve C4 is the static pressure vs. air volume characteristic curve for a fan with a frequency of 40 Hz. Curve C5 is the static pressure vs. air volume characteristic curve for a fan with a frequency of 50 Hz. Curves C6 and C6' are duct resistance curves. Coefficients B and B' of curves C6 and C6' are determined, for example, according to the predetermined length, height, and shape of the duct extending downstream of the fan.
[0078] The simulator 54 that performs a simulation using a theoretical physical model is stored in advance in the storage unit 91. The theoretical physical model is, for example, a flow rate calculation model, a pressure calculation model, an air volume calculation model, a static pressure calculation model, etc.
[0079] The storage unit 91 stores the first operating data D12, D13 and the third operating data D32, D33 generated by the generation unit 991.
[0080] The storage unit 91 stores the first control models M12 and M13 and the second control models M22 and M23 that the learning unit 992 learns.
[0081] The storage unit 91 stores the second operating data D22, D23 acquired by the correction unit 993.
[0082] (2-3-2) Input section The input unit 92 is an input device such as a keyboard, a mouse, etc. Various commands and various pieces of information can be input to the control device 9 using the input unit 92.
[0083] (2-3-3) Display section The display unit 93 is a display device such as a monitor, etc. The display unit 93 can display various data stored in the storage unit 91.
[0084] (2-3-4) Communications Department The communication unit 94 is a network interface device for communicating with the heat source side control device 19 and the use side control device 39 via the network NW.
[0085] (2-3-5) Control Unit The control unit 99 is a processor such as a CPU or a GPU. The control unit 99 reads and executes programs stored in the storage unit 91 to realize various functions of the control device 9. The control unit 99 can also write calculation results to the storage unit 91 and read information stored in the storage unit 91 according to the programs.
[0086] As shown in FIG. 5, the control unit 99 has, as functional blocks, a generation unit 991, a learning unit 992, a correction unit 993, and an operation unit 994.
[0087] (2-3-5-1) Generation part The generating unit 991 generates the first operating data D12 by a simulation using a theoretical physical model based on the local information 52. The first operating data D12 is dummy operating data of the water transport system 2.
[0088] The operating data of the water transport system 2 includes the flow rate, pressure, and frequency of each of the first pumps 15a, 15b and the second pump 22. For example, the generating unit 991 determines the water piping resistance curve in the pressure-flow rate characteristic curve diagram for each of the first pumps 15a, 15b and the second pump 22 based on the length, height, and shape of the water piping WP1 to WP5. The generating unit 991 may further determine the water piping resistance curve in the pressure-flow rate characteristic curve diagram for each of the first pumps 15a, 15b and the second pump 22 by taking into consideration the degree of mutual attraction of water between the first pump 15a and the second pump 22, the degree of mutual attraction of water between the first pump 15b and the second pump 22, and the degree of mutual attraction of water between the first pumps 15a and 15b. Then, the generation unit 991 sets the pressure-flow characteristic curve diagrams of the first pumps 15a, 15b and the second pump 22, for which the water piping resistance curves have been determined, and the target flow rates of the first pumps 15a, 15b and the second pump 22, in the simulator 54, thereby calculating the flow rates, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22 when the target flow rates of the first pumps 15a, 15b and the second pump 22 are the values set in the simulator 54. For example, the generation unit 991 changes the target flow rates of the first pumps 15a, 15b and the second pump 22 to various values and sets them in the simulator 54, thereby generating a plurality of combinations (combinations made up of 12 values) made up of the target flow rates, flow rates, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22, and sets these combinations as the first operating data D12.
[0089] The generation unit 991 generates third operating data D32 by simulation based on the corrected local information 52. The third operating data D32 is operating data of the water transport system 2. For example, the generation unit 991 generates the third operating data D32 in the same manner as when generating the first operating data D12.
[0090] The generating unit 991 generates the first operating data D13 by a simulation using a theoretical physical model based on the local information 53. The first operating data D13 is dummy operating data of the pneumatic conveying system 3.
[0091] The operating data of the air conveying system 3 includes the air volume, static pressure, and frequency of each of the supply air fan 36 and the dispersion fans 441, 442. For example, the generation unit 991 determines the duct resistance curves in the static pressure-air volume characteristic curve diagrams of each of the supply air fan 36 and the dispersion fans 441, 442 based on the length, height, and shape of the supply air duct 43. The generation unit 991 may further determine the duct resistance curves in the static pressure-air volume characteristic curve diagrams of each of the supply air fan 36 and the dispersion fans 441, 442 by taking into consideration the degree of air attraction between the supply air fan 36 and the dispersion fan 441, the degree of air attraction between the supply air fan 36 and the dispersion fan 442, and the degree of air attraction between the dispersion fans 441, 442. The generation unit 991 then sets the static pressure-airflow characteristic curve diagrams of the supply air fan 36 and the dispersion fans 441, 442, for which the duct resistance curves have been determined, and the target airflows of the supply air fan 36 and the dispersion fans 441, 442, in the simulator 54, thereby calculating the airflow, static pressure, and frequency of the supply air fan 36 and the dispersion fans 441, 442 when the target airflows of the supply air fan 36 and the dispersion fans 441, 442 are the values set in the simulator 54. For example, the generation unit 991 changes the target airflows of the supply air fan 36 and the dispersion fans 441, 442 to various values and sets them in the simulator 54, thereby generating a plurality of combinations (combinations made up of 12 values) made up of the target airflows, airflows, static pressures, and frequencies of the supply air fan 36 and the dispersion fans 441, 442, and sets these as the first operating data D13.
[0092] The generation unit 991 generates third operating data D33 by simulation based on the corrected local information 53. The third operating data D33 is operating data of the pneumatic conveying system 3. For example, the generation unit 991 generates the third operating data D33 in the same way as when generating the first operating data D13.
[0093] (2-3-5-2) Learning Department The learning unit 992 uses the first operating data D12 to learn the first control model M12. The first control model M12 is a learning model for controlling the water transportation system 2. In this embodiment, the first control model M12 is a reinforcement learning model. The learning unit 992 may learn the first control model M12 before the water transportation system 2 is installed on-site. Alternatively, the learning unit may learn the first control model M12 after the water transportation system 2 is installed on-site and before operation of the water transportation system 2 begins.
[0094] For example, the reinforcement learning state in the first control model M12 is specified by the target flow rate, flow rate, and frequency of each of the first pumps 15a, 15b and the second pump 22. The reinforcement learning action in the first control model M12 is to change one or more of the target flow rates of the first pumps 15a, 15b and the second pump 22 by a predetermined value. The reinforcement learning reward in the first control model M12 is configured to increase as the sum of the power consumption of each of the first pumps 15a, 15b calculated from the frequency and flow rate of each of the first pumps 15a, 15b, the power consumption of the second pump 22 calculated from the frequency and flow rate of the second pump 22, and the "difference between the target flow rate and the flow rate" of each of the first pumps 15a, 15b and the second pump 22 decreases. When the state is input to the first control model M12, the target flow rates of each of the first pumps 15a, 15b and the second pump 22 are calculated, which improves energy conservation and comfort.
[0095] The learning unit 992 uses the third operating data D32 to learn the second control model M22. For example, the learning unit 992 generates the second control model M22 in the same way as when learning the first control model M12.
[0096] The learning unit 992 uses the first operating data D13 to learn the first control model M13. The first control model M13 is a learning model for controlling the pneumatic conveying system 3. In this embodiment, the first control model M13 is a reinforcement learning model. The learning unit 992 may learn the first control model M13 before the pneumatic conveying system 3 is installed on-site. Alternatively, the learning unit may learn the first control model M13 after the pneumatic conveying system 3 is installed on-site and before operation of the pneumatic conveying system 3 begins.
[0097] For example, the reinforcement learning state in the first control model M13 is specified by the static pressure of the supply air fan 36 and the target airflow, air volume, and frequency of each of the dispersion fans 441 and 442. The reinforcement learning action in the first control model M13 is to change the target airflow of the supply air fan 36 by a predetermined value. The reinforcement learning reward in the first control model M13 is configured to increase as the sum of the power consumption of the supply air fan 36 calculated from the frequency and air volume of the supply air fan 36, the power consumption of each of the dispersion fans 441 and 442 calculated from the frequency and air volume of each of the dispersion fans 441 and 442, and the "difference between the target airflow and the air volume" of each of the dispersion fans 441 and 442 decreases. Therefore, when the state is input to the first control model M13, a target airflow of the supply air fan 36 that improves energy conservation and comfort is calculated.
[0098] The learning unit 992 uses the third operating data D33 to learn the second control model M23. For example, the learning unit 992 generates the second control model M23 in the same way as when learning the first control model M13.
[0099] (2-3-5-3) Correction section The correction unit 993 acquires second operating data D22 via the heat source side control device 19. The second operating data D22 is actual operating data of the water transport system 2. For example, the correction unit 993 acquires, at predetermined time intervals, the flow rates and frequencies of the first pumps 15a, 15b and the second pump 22, respectively, as the second operating data D22.
[0100] The correction unit 993 corrects the local information 52 using the second operation data D22. For example, the correction unit 993 performs a calculation by the simulator 54 that is the inverse of the calculation by the generation unit 991, and corrects the water piping resistance curves in the pressure-flow rate characteristic curve diagrams of the first pumps 15a, 15b and the second pump 22 using the second operation data D22 (for example, if the water piping resistance curve determined by the generation unit 991 is curve C3, the correction unit 993 corrects curve C3 to curve C3').
[0101] The correction unit 993 acquires second operating data D23 via the use-side control device 39. The second operating data D23 is actual operating data of the air conveying system 3. For example, the correction unit 993 acquires, at predetermined time intervals, the static pressure (measurement value of the discharge pressure sensor 78) and frequency of the air supply fan 36, and the air volumes (measurement values of the air volume sensors 771 and 772) and frequencies of each of the dispersion fans 441 and 442, as the second operating data D23.
[0102] The correction unit 993 uses the second operating data D23 to correct the local information 53. For example, the correction unit 993 uses the simulator 54 to perform a calculation that is the reverse of the calculation performed by the generation unit 991, and corrects the duct resistance curves in the static pressure-air volume characteristic curve diagrams of the supply air fan 36 and the dispersion fans 441, 442 using the second operating data D23 (for example, if the duct resistance curve determined by the generation unit 991 is curve C6, the correction unit 993 corrects curve C6 to curve C6').
[0103] (2-3-5-4) Operation section The operation unit 994 controls the water conveyance system 2 using the first control model M12 or the second control model M22. For example, the operation unit 994 inputs the target flow rate, flow rate, and frequency of each of the first pumps 15a, 15b and the second pump 22 into the first control model M12 or the second control model M22 at predetermined time intervals, and calculates the target flow rate of each of the first pumps 15a, 15b and the second pump 22 that maximizes the value. The operation unit 994 controls the flow rate of each of the first pumps 15a, 15b and the second pump 22 based on the calculated target flow rates of each of the first pumps 15a, 15b and the second pump 22.
[0104] The operation unit 994 controls the air conveying system 3 using the first control model M13 or the second control model M23. For example, the operation unit 994 inputs the static pressure of the supply air fan 36 (measured value of the discharge pressure sensor 78) and the target airflow rates (calculated according to the indoor loads of the air conditioning zones Z1 and Z2), airflow rates (measured values of the airflow sensors 771 and 772), and frequencies of the dispersion fans 441 and 442 to the first control model M13 or the second control model M23 at predetermined time intervals, and calculates the target airflow rate of the supply air fan 36 that maximizes the value. The operation unit 994 controls the airflow rate of the supply air fan 36 based on the calculated target airflow rate of the supply air fan 36.
[0105] (3) Processing (3-1) Treatment of water transport systems An example of the process of the water transport system 2 will be described with reference to the flowchart of FIG.
[0106] As shown in step S21, the control device 9 generates the first operating data D12 based on the local information 52 by performing a simulation using a theoretical physical model.
[0107] After completing step S21, as shown in step S22, the control device 9 uses the first operating data D12 to learn the first control model M12.
[0108] After completing step S22, the control device 9 controls the flow rates of the first pumps 15a and 15b and the second pump 22 using the first control model M12, as shown in step S23.
[0109] After step S23, the control device 9 acquires the second operating data D22 as shown in step S24.
[0110] After completing step S24, the control device 9 corrects the local information 52 using the second operating data D22, as shown in step S25.
[0111] After step S25, the control device 9 generates third operating data D32 by simulation based on the corrected local information 52, as shown in step S26.
[0112] After completing step S26, as shown in step S27, the control device 9 uses the third operating data D32 to learn the second control model M22.
[0113] After completing step S27, the control device 9 controls the flow rates of the first pumps 15a and 15b and the second pump 22 using the second control model M22 instead of the first control model M12, as shown in step S .
[0114] (3-2) Treatment of air conveying systems An example of the process of the air conveying system 3 will be described with reference to the flowchart of FIG.
[0115] As shown in step S31, the control device 9 generates the first operating data D13 based on the local information 53 by performing a simulation using a theoretical physical model.
[0116] After completing step S31, as shown in step S32, the control device 9 uses the first operating data D13 to learn the first control model M13.
[0117] After step S32, the control device 9 controls the air volume of the air supply fan 36 using the first control model M13, as shown in step S33.
[0118] After step S33 is completed, the control device 9 acquires the second operating data D23 as shown in step S34.
[0119] After completing step S34, the control device 9 corrects the local information 53 using the second operating data D23, as shown in step S35.
[0120] After step S35, the control device 9 generates third operating data D33 by simulation based on the corrected local information 53, as shown in step S36.
[0121] After completing step S36, as shown in step S37, the control device 9 uses the third operating data D33 to learn the second control model M23.
[0122] After step S37, as shown in step S38, the control device 9 controls the air volume of the air supply fan 36 using the second control model M23 instead of the first control model M13.
[0123] (4) Features (4-1) 2. Description of the Related Art Conventionally, there is a technique for learning a control model for controlling an equipment device by using operation data of a device similar to the equipment device when operation data of the equipment device is not available.
[0124] When the configuration of an equipment device varies significantly depending on the installation site, conventional techniques are unable to use operating data of equipment similar to the equipment device, and are therefore unable to learn a control model before the equipment device begins operation. Therefore, when the configuration of an equipment device varies significantly depending on the installation site, conventional techniques are unable to efficiently control the equipment device using a control model when the equipment device begins operation.
[0125] The control system 1 of this embodiment includes equipment 2, 3 and a control unit 99. The control unit 99 controls the equipment 2, 3. The control unit 99 generates first operating data D12, D13 based on local information 52, 53. The local information 52, 53 is information about the equipment 2, 3 at the local site where the equipment 2, 3 is installed. The first operating data D12, D13 is dummy operating data for the equipment 2, 3. The control unit 99 learns first control models M12, M13 using the first operating data D12, D13. The control unit 99 controls the equipment 2, 3 using the first control models M12, M13.
[0126] In the control system 1, the control unit 99 generates first operating data D12, D13 based on the local information 52, 53. The local information 52, 53 is information about the equipment 2, 3 at the local site where the equipment 2, 3 is installed. The first operating data D12, D13 is dummy operating data for the equipment 2, 3. Therefore, even if the configurations of the equipment 2, 3 vary significantly depending on the local site where the equipment 2, 3 is installed, the control system 1 can generate the first operating data for the equipment 2, 3 and learn the first control models M12, M13 before the equipment 2, 3 starts to operate. As a result, even if the configurations of the equipment 2, 3 vary significantly depending on the local site where the equipment 2, 3 is installed, the control system 1 can efficiently control the equipment 2, 3 using the first control models M12, M13 from the start of operation of the equipment 2, 3.
[0127] (4-2) In the control system 1 of this embodiment, the on-site information 52, 53 includes drawing information 521, 531 of the equipment 2, 3 and characteristic information 522, 532 of the equipment 2, 3. The control unit 99 generates first operating data D12, D13 based on the on-site information 52, 53 by simulation using a theoretical physics model.
[0128] As a result, the control system 1 can generate a large amount of various types of operating data.
[0129] (4-3) In the control system 1 of this embodiment, the control unit 99 acquires second operating data D22, D23. The second operating data D22, D23 is actual operating data of the equipment 2, 3. The control unit 99 corrects the local information 52, 53 using the second operating data D22, D23. The control unit 99 generates third operating data D32, D33 by simulation based on the corrected local information 52, 53. The third operating data D32, D33 is operating data of the equipment 2, 3. The control unit 99 learns second control models M22, M23 using the third operating data D32, D33. The control unit 99 controls the equipment 2, 3 using the second control models M22, M23.
[0130] The control system 1 generates third operating data D32, D33 through simulation based on the corrected local information 52, 53. The control unit 99 uses the third operating data D32, D33 to learn second control models M22, M23. The control unit 99 controls the equipment 2, 3 using the second control models M22, M23.
[0131] As a result, the control system 1 can control the facility devices 2 and 3 with higher accuracy.
[0132] (4-4) In the control system 1 of this embodiment, the facility devices 2 and 3 include first pipes WP1 to WP5 and 43 and one or more first devices 15a, 15b, 22, 36, 441, and 442. The first pipes WP1 to WP5 and 43 transport media. The first devices 15a, 15b, 22, 36, 441, and 442 control the flow of the media. The drawing information 521 and 531 include the arrangement of the first pipes WP1 to WP5 and 43 and the arrangement of the first devices 15a, 15b, 22, 36, 441, and 442. The arrangement of the first pipes WP1 to WP5 and 43 includes the length, height, and shape of the first pipes WP1 to WP5 and 43. The control unit 99 controls the first devices 15a, 15b, 22, 36, 441, and 442 using the first control models M12 and M13 or the second control models M22 and M23.
[0133] As a result, the control system 1 can automatically control one or more first devices 15a, 15b, 22, 36, 441, 442, for example, to improve energy conservation and comfort.
[0134] (4-5) In the control system 1 of this embodiment, the facility equipment 3 is an air conveying system 3. The first pipe 43 is an air supply duct 43. The medium is air. The first equipment 36, 441, 442 includes an air supply fan 36 and dispersion fans 441, 442. The operating data of the facility equipment 3 includes the air volume, static pressure, and frequency of each of the air supply fan 36 and the dispersion fans 441, 442. The control unit 99 controls the air volume of the air supply fan 36 using the first control model M13 or the second control model M23.
[0135] (4-6) In the control system 1 of this embodiment, the facility equipment 2 is a water transport system 2. The first pipes WP1 to WP5 are water piping WP1 to WP5. The medium is water. The first equipment 15a, 15b, 22 include first pumps 15a and 15b and a second pump 22. The operation data of the facility equipment 2 includes the flow rate, pressure, and frequency of each of the first pumps 15a and 15b and the second pump 22. The control unit 99 controls the flow rate of each of the first pumps 15a and 15b and the second pump 22 using the first control model M12 or the second control model M22.
[0136] As a result, the control system 1 can automatically control the flow rate balance between the first pumps 15a and 15b and the second pump 22 so as to improve, for example, energy conservation and comfort.
[0137] (5) Variations (5-1) Variation 1A In this embodiment, the reinforcement learning behavior in the first control model M12 or the second control model M22 is to change, by a predetermined value, the target flow rate of one or more of the first pumps 15a, 15b and the second pump 22. However, the reinforcement learning behavior in the first control model M12 or the second control model M22 may also be to change, by a predetermined value, the frequency of one or more of the first pumps 15a, 15b and the second pump 22.
[0138] As a result, the control system 1 can automatically control the flow rate balance between the first pumps 15a and 15b and the second pump 22 so as to improve, for example, energy conservation and comfort.
[0139] (5-2) Variation 1B In this embodiment, the reinforcement learning behavior in the first control model M13 or the second control model M23 is to change, by a predetermined value, the target air volume of the supply air fan 36. However, the reinforcement learning behavior in the first control model M13 or the second control model M23 may also be to change, by a predetermined value, the frequency of one or more of the supply air fan 36 and the dispersion fans 441, 442.
[0140] As a result, the control system 1 can automatically control the balance of the airflow rates of the supply air fan 36 and the dispersion fans 441, 442 so as to improve energy conservation and comfort, for example.
[0141] (5-3) Variation 1C In this embodiment, the generation unit 991 sets the pressure-flow characteristic curve diagrams of the first pumps 15a, 15b and the second pump 22, for which the water piping resistance curves have been determined, and the target flow rates of the first pumps 15a, 15b and the second pump 22, in the simulator 54, and thereby calculates the flow rates, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22 when the target flow rates of the first pumps 15a, 15b and the second pump 22 are the values set in the simulator 54.
[0142] However, the generation unit 991 may set, in the simulator 54, the pressure-flow characteristic curve diagrams of the first pumps 15a, 15b and the second pump 22, for which the water piping resistance curves have been determined, and the target pressures of the first pumps 15a, 15b and the second pump 22, thereby calculating the flow rates, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22 when the target pressures of the first pumps 15a, 15b and the second pump 22 are the values set in the simulator 54. For example, the generation unit 991 may change the target pressures of the first pumps 15a, 15b and the second pump 22 to various values and set them in the simulator 54 to generate a plurality of combinations (combinations made up of 12 values) made up of the flow rates, target pressures, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22, and use these as the first operation data D12.
[0143] In this case, for example, the state of reinforcement learning in the first control model M12 or the second control model M22 is specified by the target pressure, pressure, and frequency of each of the first pumps 15a, 15b and the second pump 22. The behavior of reinforcement learning in the first control model M12 or the second control model M22 is to change, by a predetermined value, one or more target pressures of the first pumps 15a, 15b and the second pump 22. The reward of reinforcement learning in the first control model M12 or the second control model M22 is configured to increase as the sum of the power consumption of each of the first pumps 15a, 15b calculated from the frequency and flow rate of each of the first pumps 15a, 15b, the power consumption of the second pump 22 calculated from the frequency and flow rate of the second pump 22, and the "difference between the target pressure and the pressure" of each of the first pumps 15a, 15b and the second pump 22 decreases. When the state is input to the first control model M12 or the second control model M22, the target pressures of the first pumps 15a, 15b and the second pump 22 that improve energy saving and comfort are calculated.
[0144] For example, the operation unit 994 inputs the target pressures, pressures, and frequencies of the first pumps 15a, 15b and the second pump 22 into the first control model M12 or the second control model M22 at predetermined time intervals, and calculates the target pressures of the first pumps 15a, 15b and the second pump 22 that maximize the value. The operation unit 994 controls the pressures of the first pumps 15a, 15b and the second pump 22 based on the calculated target pressures of the first pumps 15a, 15b and the second pump 22.
[0145] (5-4) Variation 1D In this embodiment, the generation unit 991 sets the static pressure-air volume characteristic curve diagrams of the supply air fan 36 and the dispersion fans 441, 442, for which the duct resistance curves have been determined, and the target air volumes of the supply air fan 36 and the dispersion fans 441, 442, in the simulator 54, and thereby calculates the air volumes, static pressures, and frequencies of the supply air fan 36 and the dispersion fans 441, 442 when the target air volumes of the supply air fan 36 and the dispersion fans 441, 442 are the values set in the simulator 54.
[0146] However, the generation unit 991 may set in the simulator 54 the static pressure-air volume characteristic curve diagrams of the supply air fan 36 and the dispersion fans 441, 442, for which the duct resistance curves have been determined, as well as the target discharge pressure (target static pressure) of the supply air fan 36 and the target air volumes of the dispersion fans 441, 442, and calculate the air volumes, static pressures, and frequencies of the supply air fan 36 and the dispersion fans 441, 442 when the target static pressure of the supply air fan 36 and the target air volumes of the dispersion fans 441, 442 are the values set in the simulator 54. For example, the generation unit 991 changes the target static pressure of the supply air fan 36 and the target air volume of each of the dispersion fans 441, 442 to various values and sets them in the simulator 54, thereby generating multiple combinations (combinations consisting of 12 values) consisting of the target static pressure, air volume, static pressure, and frequency of the supply air fan 36 and the target air volume, air volume, static pressure, and frequency of each of the dispersion fans 441, 442, and sets these as the first operating data D13.
[0147] In this case, for example, the state of reinforcement learning in the first control model M13 or the second control model M23 is specified by the static pressure of the supply air fan 36 and the target air volume, airflow rate, and frequency of each of the dispersion fans 441, 442. The behavior of reinforcement learning in the first control model M13 or the second control model M23 is to change the target static pressure of the supply air fan 36 by a predetermined value. The reward of reinforcement learning in the first control model M13 or the second control model M23 is configured to increase as the sum of the power consumption of the supply air fan 36 calculated from the frequency and airflow rate of the supply air fan 36, the power consumption of each of the dispersion fans 441, 442 calculated from the frequency and airflow rate of each of the dispersion fans 441, 442, the "difference between the target static pressure and static pressure" of the supply air fan 36, and the "difference between the target air volume and airflow rate" of each of the dispersion fans 441, 442 decreases. Therefore, when a state is input to the first control model M13 or the second control model M23, a target static pressure of the air supply fan 36 that improves energy conservation and comfort is calculated.
[0148] For example, the operation unit 994 inputs the static pressure of the supply air fan 36 (measured value of the discharge pressure sensor 78) and the target airflow (calculated according to the indoor load of the air conditioning zones Z1, Z2), airflow (measured value of the airflow sensors 771, 772), and frequency of each of the dispersion fans 441, 442 into the first control model M13 or the second control model M23 at predetermined time intervals, and calculates the target static pressure of the supply air fan 36 that maximizes the value. The operation unit 994 controls the static pressure of the supply air fan 36 based on the calculated target static pressure of the supply air fan 36.
[0149] (5-5) Variation 1E In this embodiment, the operating data of the water transport system 2 includes the flow rate, pressure, and frequency of each of the first pumps 15a, 15b and the second pump 22. However, the operating data of the water transport system 2 may further include the opening degree of the flow rate adjustment valve 23. As a result, the control system 1 can automatically control the opening degree of the flow rate adjustment valve 23 so as to, for example, improve energy conservation and comfort.
[0150] (5-6) Variation 1F In this embodiment, the operating data of the water transport system 2 includes the flow rate, pressure, and frequency of each of the first pumps 15a, 15b and the second pump 22. However, the operating data of the water transport system 2 may further include at least one of the measurement values of the inlet temperature sensors 81a, 81b, the measurement value of the outlet temperature sensors 82a, 82b, the measurement value of the supply water temperature sensor 83, and the measurement value of the return water temperature sensor 84. In this case, for example, the local information 52 includes the designed indoor load and outdoor air load. The generator 991 calculates at least one of the measurement values of the inlet temperature sensors 81a, 81b, the measurement value of the outlet temperature sensors 82a, 82b, the measurement value of the supply water temperature sensor 83, and the measurement value of the return water temperature sensor 84 using the simulator 54 based on the designed indoor load and outdoor air load. The generation unit 991 calculates the target air volume or target pressure for each of the supply air fan 36 and the dispersion fans 441, 442 based on at least one of the measurement values of the inlet temperature sensors 81a, 81b, the measurement values of the outlet temperature sensors 82a, 82b, the measurement value of the supply water temperature sensor 83, and the measurement value of the return water temperature sensor 84. The generation unit 991 calculates the flow rate, pressure, and frequency for each of the first pumps 15a, 15b and the second pump 22 by setting, in the simulator 54, the pressure-flow rate characteristic curve diagrams for each of the first pumps 15a, 15b and the second pump 22, for which the water piping resistance curves have been determined, and the target flow rate or target pressure for each of the first pumps 15a, 15b and the second pump 22.
[0151] As a result, the control system 1 can control the water transport system 2 with even greater accuracy by including at least one of the measurement values of the inlet temperature sensors 81a, 81b, the measurement values of the outlet temperature sensors 82a, 82b, the measurement value of the supply water temperature sensor 83, and the measurement value of the return water temperature sensor 84 in the reinforcement learning state in the first control model M12 or the second control model M22.
[0152] (5-7) Variation 1G In this embodiment, the air conveying system 3 includes the dispersion fans 441 and 442. However, the air conveying system 3 may include a damper instead of or in addition to the dispersion fans 441 and 442. As a result, by including the opening degree of the damper in the operation data of the air conveying system 3, the control system 1 can automatically control the opening degree of the damper so as to, for example, improve energy conservation and comfort.
[0153] (5-8) Variation 1H In this embodiment, the operating data of the air conveying system 3 includes the air volume, static pressure, and frequency of each of the supply air fan 36 and the dispersion fans 441, 442. However, the operating data of the air conveying system 3 may further include at least one of the measurement value of the supply air temperature sensor 71, the measurement value of the supply air humidity sensor 72, the measurement value of the return air temperature sensor 73, the measurement value of the return air humidity sensor 74, the measurement value of the indoor temperature sensors 751, 752, and the measurement value of the indoor humidity sensors 761, 762. In this case, for example, the local information 53 includes the design indoor load and outdoor air load. The generation unit 991 calculates, using the simulator 54 based on the designed indoor load and outdoor load, at least one of the measurement value of the supply air temperature sensor 71, the measurement value of the supply air humidity sensor 72, the measurement value of the return air temperature sensor 73, the measurement value of the return air humidity sensor 74, the measurement value of the indoor temperature sensors 751, 752, and the measurement value of the indoor humidity sensors 761, 762. The generation unit 991 calculates the target airflow rates of the supply air fan 36 and the distribution fans 441, 442 (or the target static pressure of the supply air fan 36 and the target airflow rates of the distribution fans 441, 442) based on at least one of the measurement value of the supply air temperature sensor 71, the measurement value of the supply air humidity sensor 72, the measurement value of the return air temperature sensor 73, the measurement value of the return air humidity sensor 74, the measurement value of the indoor temperature sensors 751, 752, and the measurement value of the indoor humidity sensors 761, 762. The generation unit 991 calculates the air volume, static pressure, and frequency of each of the supply air fan 36 and the dispersion fans 441, 442 by setting the static pressure / air volume characteristic curve diagrams of the supply air fan 36 and the dispersion fans 441, 442, for which the duct resistance curves have been determined, and the target air volumes of each of the supply air fan 36 and the dispersion fans 441, 442 (or the target static pressure of the supply air fan 36 and the target air volumes of each of the dispersion fans 441, 442) in the simulator 54.
[0154] As a result, the control system 1 can control the air conveying system 3 with even greater accuracy by including at least one of the measurement values of the supply air temperature sensor 71, the supply air humidity sensor 72, the return air temperature sensor 73, the return air humidity sensor 74, the indoor temperature sensors 751, 752, and the indoor humidity sensors 761, 762 in the reinforcement learning state in the first control model M13 or the second control model M23.
[0155] (5-9) Variation 1I In this embodiment, the drawing information 531 of the air conveying system 3 includes the arrangement of the supply air duct 43, the supply air fan 36, and the dispersion fans 441 and 442. However, the drawing information 531 of the air conveying system 3 may further include the arrangement of the outside air duct 41 and the arrangement of the outside air fan 32.
[0156] In this case, for example, the characteristic information 532 of the air conveying system 3 is a static pressure-air volume characteristic curve diagram of each of the air supply fan 36, the dispersion fans 441 and 442, and the outside air fan 32.
[0157] For example, the generation unit 991 determines a duct resistance curve in the static pressure-air volume characteristic curve diagram for each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32, based on the length, height, and shape of the supply air duct 43 and the length, height, and shape of the outdoor air duct 41. Then, the generation unit 991 sets the static pressure-air volume characteristic curve diagram for each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32, for which the duct resistance curve has been determined, and the target air volumes for each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32, in the simulator 54, thereby calculating the air volumes, static pressures, and frequencies for each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32 when the target air volumes for each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32 are the values set in the simulator 54. For example, the generation unit 991 changes the target air volume of each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32 to various values and sets them in the simulator 54, thereby generating multiple combinations (combinations consisting of 16 values) consisting of the target air volume, air volume, static pressure, and frequency of each of the supply air fan 36, the dispersion fans 441, 442, and the outdoor air fan 32, and setting these as the first operating data D13.
[0158] For example, the state of reinforcement learning in the first control model M13 or the second control model M23 is specified by the static pressure of the supply air fan 36, the target air volume, air flow rate, and frequency of each of the dispersion fans 441 and 442, and the target air volume and frequency of the outdoor air fan 31. The behavior of reinforcement learning in the first control model M13 or the second control model M23 is to change the target air volume of the supply air fan 36 by a predetermined value. The reward for reinforcement learning in the first control model M13 or the second control model M23 is configured to increase as the sum of the power consumption of supply air fan 36 calculated from the frequency and air volume of supply air fan 36, the power consumption of each of distribution fans 441, 442 calculated from the frequency and air volume of each of distribution fans 441, 442, the power consumption of outdoor air fan 31 calculated from the frequency and air volume of outdoor air fan 31, the "difference between the target air volume and the air volume" of each of distribution fans 441, 442, and the "difference between the target air volume and the air volume" of outdoor air fan 31 decreases. Therefore, when a state is input to the first control model M13 or the second control model M23, a target air volume of supply air fan 36 that improves energy conservation and comfort is calculated.
[0159] For example, the operation unit 994 inputs to the first control model M13 or the second control model M23 at predetermined time intervals the static pressure of the supply air fan 36 (measurement value of the discharge pressure sensor 78), the target airflow (calculated according to the indoor load of the air conditioning zones Z1, Z2), airflow (measurement value of the airflow sensors 771, 772) and frequency of each of the dispersion fans 441, 442, and the target airflow (calculated according to the outdoor air load estimated from the measurement value of the outdoor air temperature sensor 61 or the measurement value of the outdoor air humidity sensor 62) and frequency of the outdoor air fan 36, and calculates the target airflow of the supply air fan 36 that maximizes the value. The operation unit 994 controls the airflow of the supply air fan 36 based on the calculated target airflow of the supply air fan 36.
[0160] (5-10) 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]
[0161] 1. Control System 2. Water conveyance system (equipment) 3. Air conveying system (equipment) 15a, 15b, 22 Pump (first equipment) 36,441,442 Fan (first equipment) 43 Duct (1st pipe) 52,53 Local Information 99 Control Unit 521,531 Drawing information 522,532 characteristic information D12, D13 First operation data D22, D23 Second operation data D32, D33 Third operation data M12, M13 1st control model M22, M23 Second control model WP1~WP5 Water piping (1st pipe) [Prior art documents] [Patent documents]
[0162] [Patent Document 1] Japanese Patent Publication No. 2022-145655
Claims
1. Equipment (2, 3); A control unit (99) that controls the facility equipment; Equipped with The control unit generating first operating data (D12, D13) that is dummy operating data of the facility equipment based on local information (52, 53) that is information about the facility equipment at a site where the facility equipment is installed; learning a first control model (M12, M13) using the first operating data; Controlling the facility equipment using the first control model. Control system (1).
2. The on-site information includes drawing information (521, 531) of the facility equipment and characteristic information (522, 532) of the facility equipment, The control unit generates the first operating data by a simulation using a theoretical physical model based on the on-site information. A control system (1) according to claim 1.
3. The control unit Acquire second operation data (D22, D23) which is actual operation data of the facility device; correcting the on-site information using the second operating data; generating third operation data (D32, D33) of the facility equipment by the simulation based on the corrected on-site information; Using the third operating data, a second control model (M22, M23) is learned; Controlling the facility equipment using the second control model. A control system (1) according to claim 2.
4. The facility equipment includes a first pipe (WP1 to WP5, 43) that transports a medium, and one or more first devices (15a, 15b, 22, 36, 441, 442) that control the flow of the medium, the drawing information includes an arrangement of the first pipe and an arrangement of the first device; the configuration of the first tube includes a length, a height, and a shape of the first tube; the control unit controls the first device using the first control model or the second control model. A control system (1) according to any one of claims 1 to 3.
5. The facility equipment is a pneumatic conveying system (3), the first pipe is a duct (43); the medium is air, the first device includes a fan (36, 441, 442); The operation data of the facility equipment includes an air volume of the fan, a static pressure of the fan, and a frequency of the fan; the control unit controls the airflow rate of the fan using the first control model or the second control model. A control system (1) according to claim 4.
6. The facility equipment is a water conveyance system (2), The first pipes are water pipes (WP1 to WP5), the medium is water, the first device includes a pump (15a, 15b, 22); The operational data of the facility equipment includes a flow rate of the pump, a pressure of the pump, and a frequency of the pump; the control unit controls the flow rate of the pump using the first control model or the second control model. A control system (1) according to claim 4.
Citation Information
Patent Citations
Correcting device, predicting device, method, program, and correcting model
JP2022145655A