Control device, control system, and control method

Abstract

The present invention involves: an optimization unit (13) that calculates control data including a step target value by optimization calculation on the basis of simulation results obtained by causing a simulator unit (21) to perform a simulation; an optimal control calculation unit (14) that starts up controlled equipment (4), collects sensor data relating to a controlled variable as measured after the startup of the controlled equipment (4), and calculates a control amount for the controlled equipment (4) on the basis of the collected sensor data in order for the controlled variable to follow the step target value calculated by the optimization unit (13); and a control unit (15) that controls the controlled equipment (4) using the control amount calculated by the optimal control calculation unit (14).

Description

Control device, control system, and control method

[0001] The present disclosure relates to a control device, a control system, and a control method that control a device to be controlled (hereinafter referred to as a "controlled device") so that a controlled quantity, which is a measurement value aiming for a target value, reaches a target value at a target time (hereinafter referred to as a "target time").

[0002] Conventionally, in systems such as air conditioning systems, a technology has been known that uses modern control techniques, such as LQR (Linear-Quadratic Regulator), to calculate a control variable for a controlled device so that the controlled variable reaches a target value at a target time. Modern control techniques track the behavior of the controlled device and calculate a control variable for the controlled device based on the difference between a target value and a current value that aims for the target value. Modern control techniques are widely used to achieve stable control. Modern control is an effective method for calculating a control variable that allows the controlled variable to quickly reach a target value while minimizing the energy consumption of the controlled device. Meanwhile, as a method for calculating an optimal control input for system control while reducing the power consumption of the entire system, for example, an optimization method, such as a genetic algorithm, is known (see, for example, Patent Document 1). This method uses the power consumption of devices installed in the system as an objective function, and searches for decision variables that minimize the objective function through optimization calculations to obtain optimal setting values ​​for the devices. Optimization methods are methods for finding optimal solutions by adjusting variables under constraints, and are effective methods for achieving goals while optimizing efficiency.

[0003] Japanese Patent Application Laid-Open No. 2021-89089

[0004] Controlled equipment can sometimes be operated more efficiently by reducing its capacity to a certain extent rather than operating it at full power. The relationship between the capacity and power consumption of controlled equipment varies depending on the controlled equipment and its operating environment. Therefore, modern control techniques are effective in calculating highly efficient control variables that reduce overall energy consumption. However, modern control is based on the concept that it is sufficient for the controlled variable to reach its target value by the target time. Therefore, conventional technologies that use modern control techniques to control controlled equipment so that the controlled variable reaches its target value at the target time have the problem of potentially causing energy loss due to the controlled variable reaching its target value too early. While optimization techniques can calculate optimal operation schedules to achieve a goal, including the start-up time and various setting values ​​of the controlled equipment, optimization techniques have difficulty calculating precise control variables, i.e., continuous optimization. Therefore, control using the control variables calculated by optimization techniques results in control over a somewhat discrete time span. In this way, in a technology for controlling a controlled device so that a controlled quantity reaches a target value by a target time, even if a modern control method is directly replaced with an optimization method, another problem arises.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a control device that can perform targeted control at a target time and prevent energy loss caused by the controlled quantity reaching the target value earlier than the target time.

[0006] The control device according to the present disclosure is a control device that controls a controlled device to be controlled so that a controlled quantity, which is a measurement value aiming for a target value, reaches the target value at a target time, and includes an optimization unit that calculates control data including step target values, which are step-by-step target values ​​until the controlled quantity reaches the target value, by optimization calculation based on simulation results showing the error between the controlled quantity and the target value and power consumption, obtained by having a simulator unit that executes a simulation that simulates the operation or behavior of the controlled device run a simulation; an optimal control calculation unit that starts the controlled device, collects sensor data related to the controlled quantity measured after the controlled device is started, and calculates a control quantity for the controlled device based on the collected sensor data so that the controlled quantity follows the step target value calculated by the optimization unit; and a control unit that controls the controlled device with the control quantity calculated by the optimal control calculation unit.

[0007] According to the present disclosure, the control device can perform targeted control at a target time and prevent energy loss caused by the controlled variable reaching the target value earlier than the target time.

[0008] 1A is a diagram for explaining the advantages and disadvantages of modern control and optimization techniques. FIG. 1A is a diagram for explaining the advantages and disadvantages of a control engineering approach, in other words, a modern control technique, when controlling the fan airflow rate so that the room temperature in the room reaches a temperature set as a target value at a target time. FIG. 1B is a diagram for explaining the advantages and disadvantages of an optimization approach, in other words, an optimization technique, when controlling the fan airflow rate so that the room temperature in the room reaches a temperature set as a target value at a target time.

[0034] FIG. 1B is a diagram for explaining the advantages and disadvantages of a control system including a control device according to a first embodiment. In the first embodiment, the control device is configured to calculate a stage target value of the room room at a stage time by optimization calculation in a "pre-optimization process" and a control amount calculated by modern control in a "modern control process" to follow the stage target value calculated by optimization calculation. FIG. 3A is a diagram for explaining an example of a stage target value of the room room at a stage time calculated in the "pre-optimization process." FIG. 3B is a diagram for explaining an example of a control amount calculated to follow the stage target value calculated in the "modern control process." 6A and 6B are diagrams illustrating an example of a hardware configuration of the control device according to the first embodiment.

[0009] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. Embodiment 1 As described above, modern control and optimization techniques are different approaches, each with its own advantages and disadvantages. Here, the advantages and disadvantages of modern control and optimization techniques will be described with reference to the drawings.

[0010] FIG. 1 is a diagram for explaining the advantages and disadvantages of modern control and optimization methods. Here, as an example, an air conditioning system is used to control the room temperature so that it reaches a temperature set as a target value at a target time, and the advantages and disadvantages of modern control and optimization methods are explained. The air conditioning system here is assumed to be, for example, a chiller system that conditions air in air-conditioned rooms in a building. The chiller system is assumed to be a general chiller system, and a simple configuration example will be explained.

[0011] The chiller system includes air conditioning-related equipment, such as an air-cooled chiller, a cushion tank, an outdoor air conditioner, and an AHU (air handling unit). The air-cooled chiller and outdoor air conditioner are connected via a chilled water circuit. The air-cooled chiller is a cooling device that uses outside air to release heat. The outdoor air conditioner controls the temperature of the chilled water in the chiller system. The outdoor air conditioner includes a condenser and a fan. The AHU is an air processing device composed of a filter, damper, fan, humidifier, cooler, etc., and conditions the air for cooling or heating using chilled water whose temperature is controlled by the outdoor air conditioner, and supplies the air to living spaces. The cushion tank absorbs water pressure fluctuations in the chilled water circuit within the chiller system. The chilled water circuit is equipped with a pump for circulating the chilled water. The chilled water circuit is also equipped with a bypass valve and a valve for controlling the flow of chilled water within the chiller system. These pumps, bypass valves, and valves are also included in air conditioning-related equipment. The chiller system also includes various sensors, such as sensors that measure the water supply temperature at various points in the chilled water circuit, sensors that measure the water supply differential pressure, sensors that measure the flow rate of each bypass valve, sensors that measure the room temperature, sensors that measure the outside air temperature, and sensors that measure the power consumption within the chiller system. The air conditioning-related equipment and the various sensors are interconnected via a control network.

[0012] Here, the aim is to have the room temperature reach a target value at a target time by controlling the fan airflow rate in the AHU. In an air conditioning system, air conditioning-related equipment is the equipment that is the target of control (hereinafter referred to as "controlled equipment"). Hereinafter, the fan airflow rate in the AHU will be simply referred to as the fan airflow rate. As described above, in this example, the fan airflow rate is the controlled variable. Furthermore, the room temperature is controlled to approach the target value by controlling the fan airflow rate. The room temperature can also be said to be a controlled variable, which is controlled by controlling the controlled variable (here, the fan airflow rate). In the following description, when the term "controlled variable" is used, the "controlled variable" refers to some measured value that is aimed to reach a target value at a target time.

[0013] FIG. 1A is a diagram illustrating advantages and disadvantages of using a control engineering approach, in other words, a modern control method, to control the fan airflow rate so that the room temperature in a room reaches a temperature set as a target value at a target time. An example of a modern control method is the Linear-Quadratic Regulator (LQR). Since LQR is a well-known method, detailed explanation will be omitted. FIG. 1B is a diagram illustrating advantages and disadvantages of using an optimization approach, in other words, an optimization method, to control the fan airflow rate so that the room temperature in a room reaches a temperature set as a target value at a target time. Examples of optimization methods include genetic algorithms and Bayesian optimization. Since genetic algorithms and Bayesian optimization are well-known methods, detailed explanation will be omitted.

[0014] As shown in Figure 1A, for example, if a control engineering approach is used to control the fan airflow rate so that the room temperature reaches a target temperature at a target time, it is possible to calculate the fan airflow rate that will quickly reach the target temperature while minimizing the energy consumption of the controlled device (here, the fan). However, the control engineering approach is based on the concept that it is sufficient for some measured value (i.e., the controlled variable) to reach the target value at the target time. Therefore, the control engineering approach tends to increase the initial fan airflow rate after the controlled device in the air conditioning system starts up in order to quickly reach the target value. As a result, the room temperature may reach the target value earlier than the target time, resulting in energy loss, i.e., energy loss from the energy consumed by the fan or other controlled devices, and energy loss from the energy required to maintain the room temperature until the target time if the target value is reached earlier than the target time. It is difficult to perform control that also optimizes the startup time of the controlled device using the control engineering approach. The control engineering approach also has the problem that, when it is necessary to calculate other set values ​​related to the controlled variable in the controlled device, it is difficult to calculate the optimal set value simultaneously with the controlled variable. In this example, it is difficult to calculate the set value related to the fan airflow, such as the bypass valve opening or water temperature, simultaneously with the fan airflow. Furthermore, the control engineering approach also has the problem that it is difficult to calculate the controlled variable that minimizes energy consumption when controlling a complex system.

[0015] On the other hand, as shown in Figure 1B, if the optimization approach controls the fan airflow rate so that the room temperature reaches the target temperature at the target time, an optimal operation schedule can be calculated, including the start-up time and fan airflow rate of the controlled device, so that the room temperature reaches the target value at the target time. The optimization approach can calculate optimal control variables and other setting values ​​even for complex systems. However, the optimization approach has the problem of controlling the system over a somewhat discrete time span because continuous optimization is difficult. For example, as shown in Figure 1B, the optimization approach controls the fan airflow rate for intervals separated by regular intervals. As a result, the room temperature changes stepwise at regular intervals. Thus, the optimization approach makes it difficult to continuously optimize the fan airflow rate, or in other words, to precisely calculate the fan airflow rate. In other words, the optimization approach makes it difficult to precisely control the room temperature. This is because continuously finding the optimal solution results in a large search space and a large amount of calculation, which requires a long time to find the optimal solution. Such control over a somewhat discrete time span is not desirable for controlling room temperature. As shown in Figure 1A, the control engineering approach allows for precise control of room temperature. Furthermore, the optimization approach has the problem that when an optimal operation schedule is calculated using a simulator, differences between the simulator and the actual equipment being controlled can degrade the accuracy of the calculated operation schedule.

[0016] As described above, although modern control can achieve precise control of room temperature, which is desirable, there is a problem that the room temperature may reach the target value earlier than the target time, resulting in a corresponding energy loss. In contrast, optimization methods can perform control including optimizing the startup times or various setting values ​​of the controlled devices, but they cannot perform precise adjustment of the room temperature.

[0017] The present disclosure focuses on the advantages and disadvantages of modern control and optimization methods as described above, and by combining modern control with optimization methods, it makes possible optimization of the startup of controlled equipment, including optimization of the startup time and various setting values ​​of the controlled equipment, which was difficult to achieve using modern control alone. This enables targeted control (more specifically, detailed control) to be performed at the target time, and prevents energy loss due to the controlled quantity reaching the target value earlier than expected.

[0018] FIG. 2 is a diagram illustrating an example configuration of a control system 100 including a control device 1 according to a first embodiment. In the following first embodiment, as an example, the control system 100 is assumed to be an air conditioning system such as that described above with reference to FIG. 1. Note that the air conditioning-related devices in the air conditioning system are collectively referred to as controlled devices 4. The air conditioning-related devices are subject to control by the control device 1. In the following first embodiment, as an example, the controlled variable is the fan airflow rate, and the controlled variable is the room temperature. In the following first embodiment, as an example, the control device 1 controls the fan airflow rate so that the room temperature reaches a target value at a target time. More specifically, the control device 1 calculates the fan airflow rate so that the room room temperature reaches the target value at the target time, and controls the fan in the AHU to achieve the calculated fan airflow rate. Note that the room is not illustrated in FIG. 2. The target time and the target value to be reached at the target time are determined in advance by an administrator or the like. In the first embodiment, the target value that aims to be reached at the target time is also referred to as the "final target value."

[0019] 2 , the control system 100 is configured, for example, with a control device 1, a simulator 2, a model creation device 3, and a control target device 4. Note that this is merely an example. For example, the model creation device 3 may be provided in a system other than the control system 100, and the control system 100 may be configured without the model creation device 3.

[0020] The control device 1 calculates a control amount (here, fan airflow rate) for the controlled device 4 so that a certain measurement value (i.e., controlled amount, here, room temperature) measured in the control system 100 reaches a target value at a target time, and controls the controlled device 4 to achieve the calculated control amount. The control device 1 is assumed to be mounted on, for example, a server (not shown).

[0021] The control device 1 includes a simulation execution instruction unit 11 , a simulation result collection unit 12 , an optimization unit 13 , an optimal control calculation unit 14 , a control unit 15 , and a data collection unit 16 .

[0022] The control device 1 causes the simulator 2 to execute a simulation, and based on the results of the simulation, calculates, through optimization calculation, the target value (hereinafter referred to as the "step target value") of the controlled quantity (here, room temperature) at each step time (hereinafter referred to as the "step time") up to the target time, the start time of the controlled device 4, and the set value (e.g., water supply temperature, etc.).

[0023] In embodiment 1, the process performed by the control device 1, in which the simulator 2 executes a simulation and the step target value of the controlled variable at each step time up to the target time, the start time of the controlled device 4, and the setting value are calculated by optimization calculation, is called the "pre-optimization process."

[0024] Furthermore, the control device 1 starts the controlled device 4 at the start time of the controlled device 4 calculated by performing the "pre-optimization process" using the set value (e.g., water supply temperature, etc.) calculated by performing the "pre-optimization process." While receiving feedback from the controlled device 4, the control device 1 collects data (hereinafter referred to as "sensor data") related to the controlled variable (here, room temperature) measured after the controlled device 4 is started from a sensor (not shown), and calculates a control variable (here, fan airflow) so that the controlled variable follows the stage target value calculated by performing the "pre-optimization process." The control device 1 then controls the controlled device 4, which is controlled by the calculated control variable, using the calculated control variable. Here, the control device 1 controls the fan in the AHU so that the calculated fan airflow is achieved. The control device 1 calculates the above control variable using modern control.

[0025] In embodiment 1, the process performed by the control device 1 to start the controlled device 4 based on the start-up time of the controlled device 4 calculated by performing the "pre-optimization process" and the setting value, calculate a control quantity so as to follow the stage target value of the controlled quantity at the stage time up to the target time calculated by performing the "pre-optimization process" while receiving feedback from the controlled device 4, and control the controlled device 4 with the calculated control quantity is referred to as "modern control process."

[0026] In the "pre-optimization process", a simulation execution instruction unit 11, a simulation result collection unit 12, and an optimization unit 13 function. In the "modern control process", an optimal control calculation unit 14, a control unit 15, and a data collection unit 16 function. Below, an example configuration of the control device 1 will be described while providing a detailed description of the "pre-optimization process" and "modern control process".

[0027] The simulation execution instruction unit 11 inputs operating condition parameters to the simulator 2 and causes the simulator 2 to execute a simulation. The operating condition parameters input by the simulation execution instruction unit 11 to the simulator 2 include the start-up time of the controlled device 4, the stage target value of the controlled variable (here, room temperature) at the stage time up to the target time, and the setting value of the controlled device 4.

[0028] Here, the simulator 2 will be described. The simulator 2 is a simulator that uses a general simulation technique. The simulator 2 includes a simulator unit 21 and a data storage unit 22. The simulator unit 21 simulates the operation or behavior of the control-target device 4 based on the loaded operating condition parameters, and performs a simulation of changes in the indoor environment or the energy consumption of the control-target device 4. In the first embodiment, the simulator unit 21 performs the simulation using PID control (Proportional-Integral-Derivative control), a well-known control technique. By performing the simulation using PID control, the simulator unit 21, or more specifically, the control device 1 that causes the simulator unit 21 to perform the simulation, does not need to collect data in advance to perform the simulation. The operating condition parameters loaded by the simulator unit 21 are the operating condition parameters input by the simulation execution instruction unit 11 described above.

[0029] When the simulator unit 21 performs a simulation, it calculates the error between the target room temperature and the current room temperature in the simulation environment, as well as the power consumption of the controlled device 4. The simulator unit 21 outputs data indicating the calculated error and power consumption (hereinafter referred to as "simulation results") to the control device 1. The simulator unit 21 also stores various data collected as a result of the simulation in the data storage unit 22 as simulation data. The simulation data includes operating data (data indicating the operating status, such as the control amount of the controlled device 4), sensor data (measurements measured by various sensors, such as water temperature, valve opening, power consumption, room temperature, and fan airflow), the simulation results calculated by the simulator unit 21, and operating condition parameters read by the simulator unit 21. Note that, although the simulator 2 is assumed to include the data storage unit 22 here, this is merely an example. The data storage unit 22 may be provided outside the simulator 2 in a location accessible by the simulator 2 and the model creation device 3.

[0030] Returning to the description of the control device 1, the simulation result collection unit 12 collects the simulation results from the simulator unit 21.

[0031] The optimization unit 13 uses a known optimization algorithm based on the simulation results collected by the simulation result collection unit 12 to calculate the optimal start-up time of the controlled-target device 4, the optimal stage target value of the controlled variable (here, room temperature) at the optimal stage time until the target time, and the optimal setting value of the controlled-target device 4. Known optimization algorithms include genetic algorithms and Bayesian optimization. The optimization unit 13 calculates the optimal start-up time of the controlled-target device 4, the optimal stage target value of the controlled variable at the optimal stage time until the target time, and the optimal setting value of the controlled-target device 4 by, for example, selecting a combination of the start-up time of the controlled-target device 4, the stage target value of the controlled variable at the optimal stage time until the target time, and the setting value of the controlled-target device 4 that consumes the least power.

[0032] The optimization unit 13 changes the operating condition parameters until an optimal solution is found, inputs the changed operating condition parameters to the simulator 2 via the simulation execution instructing unit 11, collects simulation results via the simulation result collecting unit 12, and repeats the optimization calculation as described above to search for an optimal solution. The optimization unit 13 may appropriately set the startup time of the controlled-target device 4, a set of stage time and stage target value at the stage time, and the setting value of the controlled-target device 4 for the operating condition parameters when first instructing the simulator unit 21 to execute a simulation. For example, initial values ​​may be set in advance, and the optimization unit 13 may set the startup time of the controlled-target device 4, the set of stage time and stage target value, and the setting value of the controlled-target device 4 to the initial values. Alternatively, setting conditions may be defined in advance, and the optimization unit 13 may calculate the startup time of the controlled-target device 4, the set of stage time and stage target value, and the setting value of the controlled-target device 4 in accordance with the setting conditions.

[0033] The functions of the simulation execution instructing unit 11 and the simulation result collecting unit 12 may be provided by the optimization unit 13. In this case, the control device 1 may be configured without the simulation execution instructing unit 11 and the simulation result collecting unit 12.

[0034] When the optimization unit 13 finds an optimal solution, it outputs data (hereinafter referred to as "control data") including the optimal start-up time of the controlled device 4, the optimal step target value of the controlled variable at the optimal step time up to the target time, and the optimal setting value of the controlled device 4 to the optimal control calculation unit 14.

[0035] For example, the optimization unit 13 calculates the start time of the controlled-target device 4, the stage target values ​​of the controlled variables at the stage times up to the target time, and the set value of the controlled-target device 4 as follows. Note that, hereinafter, the set values ​​are, by way of example, the bypass valve opening and the supply water temperature. {Start time} = {6:50} {Set value} = [{Bypass valve opening, 50%}, {Supply water temperature, 40°C}] {Time, target value} = [{7:00, 22°C}, {7:15, 23°C}] In this way, the optimization unit 13 calculates the stage target values ​​of the controlled variables at the stage times up to the target time as pairs of associated stage times and stage target values. The optimization unit 13 can calculate multiple pairs of stage times and stage target values. It is sufficient that the optimization unit 13 calculates one or more pairs of stage times and stage target values.

[0036] The optimal control calculation unit 14 starts up the controlled device 4 at the optimal start-up time calculated by the optimization unit 13, with the optimal setting values ​​calculated by the optimization unit 13, based on the control data output from the optimization unit 13. Note that the controlled device 4 started up by the optimal control calculation unit 14 here is an actual device. When the optimal control calculation unit 14 starts up the controlled device 4 and the controlled device 4 starts up, the data collection unit 16 collects sensor data related to the controlled variable (here, room temperature). The data collection unit 16 collects the sensor data at a preset cycle while the controlled device 4 is operating.

[0037] Based on the sensor data collected by the data collection unit 16, the optimal control calculation unit 14 uses a known modern control algorithm to calculate a controlled variable (here, the fan air volume) so as to follow the optimal stage target value at the optimal stage time calculated by the optimization unit 13. The known modern control algorithm is an algorithm such as LQR. The optimal control calculation unit 14 calculates the controlled variable at a predetermined cycle. The optimal control calculation unit 14 calculates the controlled variable using a model (hereinafter referred to as a "system identified model") that models the relationship between the controlled variable and power consumption and the relationship between the controlled variable and the controlled variable. Modern control is a method that enables efficient control taking energy consumption into account, but requires system identification of power consumption. In other words, the above-mentioned system identified model must be prepared. The system identified model is created by the model creation device 3 before the "modern control process" is executed by the control device 1. Details of the model creation device 3 will be described later.

[0038] The optimal control calculation unit 14 determines, based on the sensor data collected by the data collection unit 16, which stage target value should be followed among the stage target values ​​set in the control data so as to follow the stage target value corresponding to the current time. The optimal control calculation unit 14 calculates a control amount using a system identified model so as to follow the stage target value determined to be followed. The optimal control calculation unit 14 can determine the stage target value to be followed by comparing the current time with the stage time associated with the stage target value in the control data. The optimal control calculation unit 14 calculates a control amount so as to follow the optimal stage target value while receiving feedback from the controlled device 4.

[0039] The function of the data collection unit 16 may be provided by the optimal control calculation unit 14. In this case, the control device 1 may be configured without the data collection unit 16.

[0040] The control unit 15 operates the control target device 4 (here, the fan of the AHU) that is controlled using the control amount (here, the fan air volume) calculated by the optimal control calculation unit 14.

[0041] Here, an image of the stage target value of the room temperature at a stage time calculated by optimization calculation in the "pre-optimization process" in the control device 1 and the control amount calculated by modern control in the "modern control process" to follow the stage target value calculated by optimization calculation will be described using drawings. FIG. 3 is a diagram for explaining an image of the stage target value of the room room at a stage time calculated by optimization calculation in the "pre-optimization process" in the control device 1 and the control amount calculated by modern control in the "modern control process" to follow the stage target value calculated by optimization calculation in the "modern control process" in embodiment 1. FIG. 3A is a diagram for explaining an image of an example of the stage target value of the room room at a stage time calculated in the "pre-optimization process," and FIG. 3B is a diagram for explaining an image of an example of the control amount calculated to follow the stage target value calculated in the "modern control process." In FIG. 3, it is assumed that the target time is set to "7:30" and the final target value is set to "24°C." Furthermore, the control device 1 calculates the control amount every minute.

[0042] For example, in the control device 1, the optimization unit 13 sets the optimal start-up time of the controlled device 4 to "6:50" and calculates two optimal pairs of stage times and stage target values: one pair with the stage time being "7:00" and the stage target value being "22°C", and the other pair with the stage time being "7:15" and the stage target value being "23°C" (see Figure 3A). In this case, the optimal control calculation unit 14 starts the controlled device 4 at 6:50, calculates the control amount, in this case the fan airflow rate, every minute from 6:50 to 7:00 so as to follow the stage target value of 22°C, calculates the fan airflow rate every minute from 7:00 to 7:15 so as to follow the stage target value of 23°C, and calculates the fan airflow rate every minute from 7:15 to 7:30 so as to follow the final target value of 24°C (see Figure 3B).

[0043] In this manner, in the control device 1, the optimization unit 13 uses an optimization algorithm to calculate a pair of the startup time of the controlled device 4 and the optimal step target value at the step time until the controlled variable reaches the final target value at the target time. Then, the optimal control calculation unit 14 uses a modern control algorithm to calculate the control variable at each step up to each step time calculated by the optimization unit 13 so that the controlled variable reaches the corresponding step target value and achieves optimal energy (low power consumption). This allows the control device 1 to achieve energy-saving control, including the startup time. In other words, the control device 1 calculates the control variable in stages following the step target values ​​set in stages up to the final target value, thereby preventing the control variable from becoming too large when the controlled device 4 is started and preventing the controlled variable from reaching the final target value too early. As a result, the control device 1 can prevent energy loss caused by the controlled variable reaching the final target value too early.

[0044] The model creation device 3 creates a system-identified model to be used by the optimal control calculation unit 14 of the control device 1 in the "modern control process." It is assumed that the model creation device 3 is mounted on, for example, a server (not shown). The model creation device 3 includes a model creation unit 31 and a model storage unit 32. The model creation unit 31 creates a system-identified model using simulation data stored in the data storage unit 22 and output when the simulator unit 21 executes a simulation in the "pre-optimization process." The model creation unit 31 stores the created system-identified model in the model storage unit 32. Note that, although the model creation device 3 is illustrated as including the model storage unit 32 here, this is merely an example. The model storage unit 32 may be provided in a location outside the model creation device 3 that can be referenced by the model creation device 3 and the control device 1. In the first embodiment, the process performed by the model creation device 3 to create a system-identified model using simulation data output when the "pre-optimization process" is executed by the control device 1 is referred to as a "system identification process."

[0045] The system identified model created by the model creating unit 31 is, for example, the following model. x: state variable y: output (controlled variable) u: input (controlled variable) Here, the state variables are the bypass valve opening, water temperature, power consumption, etc. The output is the room temperature. The input is the fan airflow rate.

[0046] As described above, in the control device 1, the optimization unit 13 causes the simulator unit 21 of the simulator 2 to execute a simulation during the optimization calculation process. The simulator unit 21 performs PID control. Therefore, various data resulting from changes in the controlled variables during the optimization calculation process is output as simulation data and stored in the data storage unit 22. The model creation unit 31 creates a system-identified model using the simulation data, thereby shortening the data collection period required for creating the system-identified model. The control device 1 utilizes the simulation data obtained during the optimization process for modern control, thereby shortening the data collection period required for modern control.

[0047] The operation of the control system 100 according to the first embodiment will be described. FIG. 4 is a flowchart for describing the operation of the control system 100 according to the first embodiment. In FIG. 4, the processes of steps ST10 and ST30 are performed by the control device 1, and the process of step ST20 is performed by the model creation device 3. The control system 100 starts the operation shown in the flowchart of FIG. 4 when, for example, the control device 1 receives an instruction to perform control such that the controlled variable reaches a target value (final target value) by the target time. For example, an administrator or the like inputs an instruction to perform the above control from an administration terminal such as a PC (Personal Computer). At this time, the administrator or the like may also input the target time and the final target value. The administration terminal is connected to the control system 100 via a network. The control system 100 performs, for example, the processes of steps ST10 and ST20 once after starting operation. The control system 100 repeats the process of step ST30, for example, until the control device 1 receives a control end instruction. For example, an administrator or the like inputs the control end instruction from a management terminal. For convenience, the flowchart in Figure 4 shows the processes from step ST10 to step ST30 as a single flow.

[0048] The control device 1 performs a "pre-optimization process" (step ST10). Specifically, the control device 1 causes the simulator 2 to execute a simulation, and calculates, by optimization calculation, the stage target value of the controlled variable (here, room temperature) at each stage time up to the target time, the start time of the control-target device 4, and a set value (e.g., water supply temperature).

[0049] The model creation device 3 performs a "system identification process" (step ST20). Specifically, the model creation device 3 creates a system-identified model using simulation data output when the control device 1 executes the "pre-optimization process" in step ST10 and the simulator unit 21 executes the simulation. The model creation device 3 then stores the created system-identified model in the model storage unit 32.

[0050] The control device 1 performs a "modern control process" (step ST30). Specifically, the control device 1 starts up the control-target device 4 based on the start-up time and setting value of the control-target device 4 calculated by performing the "pre-optimization process" in step ST10, calculates a control amount (here, the fan airflow rate) while receiving feedback from the control-target device 4 so as to follow the stage target value of the controlled amount at the stage time up to the target time calculated by performing the "pre-optimization process", and controls the control-target device 4 (here, the fan of the AHU) with the calculated control amount.

[0051] In this way, based on the simulation results obtained by running a simulation on the simulator 2, the control system 100 calculates control data including stepwise target values ​​(stepwise target values) until the controlled variable reaches the target value (final target value)—more specifically, the startup time, set value, and stepwise target value at the stepwise time—by optimization calculation. The control system 100 then starts the controlled device 4 at the startup time and set value calculated by the optimization calculation, calculates the control variable using a modern control technique so that the controlled variable follows the stepwise target value, and controls the controlled device 4. By calculating the stepwise target value using an optimization technique, the control system 100 can control the controlled variable so that it reaches the target value (final target value) by the target time and prevent energy loss due to the controlled variable reaching the target value (final target value) earlier than expected. Furthermore, the control system 100 enables optimization of the startup of the controlled device 4, including the startup time and set value of the controlled device 4, which was difficult using modern control alone. In modern control, if the hyperparameters (e.g., the weight matrix of the evaluation function in modern control) are inappropriate, overshooting or hunting may occur. In response to this, the control system 100 can reduce the possibility of overshooting or hunting by calculating the controlled variable so that it reaches the target value (step target value) in stages.

[0052] As described above, modern control is a method that enables efficient control taking energy consumption into account, but requires system identification of power consumption. However, in reality, accurate system identification can be difficult in complex systems involving multiple devices. For example, an air conditioning system can be considered a complex system involving multiple devices. In contrast, the control system 100 calculates a pair of target times (step times) and target values ​​(step target values) that achieve a certain degree of energy conservation through optimization calculations in the control device 1, which do not require modeling through system identification. The control system 100 then calculates controlled variables to track the step target values. Therefore, the control system 100 can achieve energy-saving control even when accurate system identification is difficult. More specifically, in the control system 100, the model creation device 3 creates a system-identified model used to calculate controlled variables that reduce power consumption in modern control using simulation data obtained during the optimization process in the control device 1. This allows the control system 100 to utilize data obtained during the optimization process (simulation data) in modern control. The control system 100 can contribute to shortening the data collection period required for modern control, more specifically, for creating a system identified model required for modern control.

[0053] The operation of the control device 1 according to the first embodiment will be described. FIG. 5 is a flowchart for describing the operation of the control device 1 according to the first embodiment. The control device 1 starts the operation shown in the flowchart of FIG. 5, for example, when it receives an instruction to perform control so that the controlled variable reaches a target value (final target value) by a target time. The control device 1 performs, for example, the processing of steps ST101 to ST103 once after starting the operation. The control device 1 repeats the processing of steps ST302 to ST303, for example, until it receives an instruction to end control. For convenience, the flowchart of FIG. 5 shows the processing of steps ST101 to ST303 as a series of steps. Regarding the processing shown in the flowchart of FIG. 5, the operations of steps ST101 to ST103 are operations executed in a "pre-optimization process," and the operations of steps ST301 to ST303 are operations executed in a "modern control process."

[0054] The simulation execution instruction unit 11 inputs operating condition parameters to the simulator 2 and causes the simulator 2 to execute a simulation (step ST101). In the simulator 2, the simulator unit 21 uses PID control to simulate the operation or behavior of the control-target equipment 4 based on the read operating condition parameters, and simulates changes in the indoor environment or the energy consumption of the control-target equipment 4. The simulator unit 21 outputs the simulation results to the control device 1. The simulator unit 21 also stores the simulation data in the data storage unit 22.

[0055] The simulation result collection unit 12 collects the simulation results from the simulator unit 21 (step ST102).

[0056] Based on the simulation results collected by the simulation result collection unit 12 in step ST102, the optimization unit 13 uses a known optimization algorithm to calculate the optimal start-up time of the controlled device 4, the optimal stage target value of the controlled variable (here, room temperature) at the optimal stage time until the target time, and the optimal setting value of the controlled device 4 (step ST103). When the optimization unit 13 finds the optimal solution, it outputs control data to the optimal control calculation unit 14.

[0057] The optimization unit 13 changes the operating condition parameters until an optimal solution is found, inputs the changed operating condition parameters to the simulator 2 via the simulation execution instruction unit 11, collects simulation results via the simulation result collection unit 12, and repeats the optimization calculation as described above to search for an optimal solution. In other words, the control device 1 changes the operating condition parameters and repeats the processing of steps ST101 to ST103 until the optimization unit 13 finds the optimal solution.

[0058] The optimum control calculation unit 14 starts up the controlled device 4 with the optimum setting value at the optimum start time calculated by the optimization unit 13 in step ST103 (step ST301). Note that the controlled device 4 started up by the optimum control calculation unit 14 here is a real device.

[0059] The data collection unit 16 collects sensor data related to the controlled variable (here, room temperature). Then, the optimal control calculation unit 14 uses the system identified model based on the sensor data collected by the data collection unit 16 and a known modern control algorithm to calculate the controlled variable (here, fan airflow rate) so as to follow the optimal stage target value at the optimal stage time calculated by the optimization unit 13 in step ST103 (step ST302).

[0060] The control unit 15 operates the control target device 4 (here, the fan of the AHU) that is to be controlled using the control amount calculated by the optimal control calculation unit 14 in step ST302 (step ST303).

[0061] The control device 1 repeats the processing of steps ST302 to ST303, determines the stage target value to be followed by comparing the current time with the stage time corresponding to the stage target value, and calculates the control amount so as to follow the determined stage target value.

[0062] In this way, the control device 1 calculates control data, including step target values, which are step-by-step target values ​​until the controlled variable reaches the target value (final target value), by optimization calculation based on the simulation results, obtained by causing the simulator unit 21 to execute the simulation and indicating the error between the controlled variable and the target value (final target value) and the power consumption. After performing the optimization calculation, the control device 1 starts the controlled device 4, collects sensor data related to the controlled variable measured after the controlled device 4 is started, and calculates the control variable for the controlled device 4 based on the collected sensor data so that the controlled variable follows the calculated step target value. The control device 1 then controls the controlled device using the calculated control variable. By calculating the step target value using an optimization technique, the control device 1 can control the controlled variable so that it reaches the target value (final target value) by the target time and prevent energy loss due to the controlled variable reaching the target value (final target value) earlier than expected. Furthermore, the control device 1 enables optimization of the startup of the controlled device 4, including the startup time and setting values ​​of the controlled device 4, which was difficult with modern control alone. In modern control, if the hyperparameters are inappropriate, overshooting or hunting may occur. In response to this, the control device 1 can reduce the possibility of overshooting or hunting by calculating the controlled variable so that it reaches the target value (step target value) in stages.

[0063] As described above, modern control is a method that enables efficient control that takes energy consumption into account, but requires system identification of power consumption. However, in reality, accurate system identification can be difficult in complex systems involving multiple devices. For example, an air conditioning system can be considered a complex system involving multiple devices. In response to this, the control device 1 calculates a pair of target times (step times) and target values ​​(step target values) that achieve a certain degree of energy conservation through optimization calculations that do not require modeling through system identification. The control device 1 then calculates controlled variables to track the step target values. Therefore, the control device 1, and the control system 100, can achieve energy-saving control even when accurate system identification is difficult. Furthermore, the control device 1 uses a system-identified model created using simulation data obtained during the optimization process to calculate controlled variables that reduce power consumption in modern control. The control device 1 can utilize data (simulation data) obtained during the optimization process for modern control. The control device 1 can contribute to shortening the data collection period required for modern control, more specifically, for creating a system-identified model required for modern control.

[0064] 6A and 6B are diagrams illustrating an example of a hardware configuration of the control device 1 according to the first embodiment. In the first embodiment, the functions of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16 are realized by a processing circuit 1001. That is, the control device 1 includes a processing circuit 1001 for calculating a control variable using a modern control technique and controlling the controlled device 4 so that the controlled variable follows a step-by-step target value calculated using an optimization technique. The processing circuit 1001 may be dedicated hardware as shown in FIG. 6A, or a processor 1004 that executes a program stored in memory as shown in FIG. 6B.

[0065] If the processing circuit 1001 is dedicated hardware, the processing circuit 1001 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof.

[0066] When the processing circuit is a processor 1004, the functions of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16 are realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 1005. The processor 1004 reads and executes the program stored in the memory 1005 to perform the functions of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16. In other words, the control device 1 includes a memory 1005 for storing a program that, when executed by the processor 1004, results in the execution of steps ST101 to ST303 of FIG. 5 described above. In addition, the program stored in memory 1005 can also be said to cause the computer to execute the processing procedures or methods of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16. Here, the memory 1005 may be, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable Read Only Memory), or an EEPROM (Electrically Erasable Programmable Read-Only Memory), a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, or a DVD (Digital Versatile Disc).

[0067] The functions of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16 may be partially implemented by dedicated hardware and partially implemented by software or firmware. For example, the control unit 15 may be implemented by a processing circuit 1001 as dedicated hardware, and the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, and the data collection unit 16 may be implemented by a processor 1004 reading and executing programs stored in a memory 1005. The control device 1 also includes an input interface device 1002 and an output interface device 1003 that communicate with devices such as the simulator 2 and the model creation device 3 via wired or wireless communication.

[0068] An example of the hardware configuration of the model creation device 3 according to the first embodiment is also configured as shown in Figures 6A and 6B. In the first embodiment, the function of the model creation unit 31 is realized by a processing circuit 1001. That is, the model creation device 3 includes a processing circuit 1001 for performing control to create a system-identified model using simulation data output when the "pre-optimization process" is executed by the control device 1. The processing circuit 1001 may be dedicated hardware as shown in Figure 6A, or a processor 1004 that executes a program stored in memory as shown in Figure 6B.

[0069] When the processing circuit is a processor 1004, the functions of the model creation unit 31 are realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in memory 1005. The processor 1004 executes the functions of the model creation unit 31 by reading and executing the program stored in memory 1005. That is, the model creation device 3 includes memory 1005 for storing a program that, when executed by the processor 1004, results in the execution of step ST20 in FIG. 4 described above. It can also be said that the program stored in memory 1005 causes a computer to execute the processing procedure or method of the model creation unit 31.

[0070] The model storage unit 32 is configured with a memory 1005, RAM, ROM, etc. The model creation device 3 also includes an input interface device 1002 and an output interface device 1003 that perform wired or wireless communication with devices such as the control device 1 and simulator 2.

[0071] In the first embodiment, the optimization unit 13 of the control device 1 may select multi-objective optimization, which allows a trade-off between comfort, energy conservation, or power consumption, in which a measured value and a target value (final target value) match. For example, a user can select a trade-off between comfort, i.e., reaching the target value (final target value) at the target time, energy conservation, and the demand value, i.e., maximum power consumption, based on whether the Pareto frontier is crossed. This allows the control system 100 to be a system that can set target values ​​for controlled variables by prioritizing energy conservation or the demand value, even if it sacrifices comfort to some extent, i.e., even if the target value (final target value) is not exactly reached at the target time. Furthermore, the control device 1 can be a device that can set target values ​​for controlled variables by prioritizing energy conservation or the demand value, even if it sacrifices comfort to some extent, i.e., even if the target value (final target value) is not exactly reached at the target time.

[0072] Furthermore, in the above-described first embodiment, the controlled variable is the room temperature and the controlled variable is the fan air volume, but this is merely an example. For example, the controlled variable may be the supply air temperature of the AHU. Any measured value measured within the air conditioning system can be the controlled variable. Also, for example, the controlled variable may be the set value of the supply air temperature of the AHU fan, or the supply air temperature of the AHU. In an air conditioning system, a controlled variable can be a value that can be controlled to control some target set value (controlled variable).

[0073] In the first embodiment described above, the control system 100 is assumed to be an air conditioning system, more specifically, a chiller system. However, this is merely an example. When the control system 100 is an air conditioning system, the air conditioning system may be a VRF (Variable Refrigerant Flow) system. A VRF system, for example, conditions the air of rooms in a building. In a VRF system, the air conditioning-related equipment that serves as the controlled device 4 includes, for example, a compressor and an indoor unit. The indoor unit includes a fan, a heat exchanger, and the like. The compressor compresses a refrigerant to a high pressure and uses it as a fluid. The refrigerant is sent from the compressor to the indoor unit, where the fan controls the flow of the refrigerant to blow air. The rotation of the fan heats or cools the air, thereby adjusting the room temperature. In a VRF system, for example, the control device 1 can control the compressor so that the room temperature reaches a target value (final target value) at a target time, with the room temperature as a controlled variable and the compressor frequency as a controlled variable. The control device 1 executes a simulation and uses an optimization method to calculate the optimal start-up time, optimal setting value, and optimal stage target value of the room temperature at the optimal stage time for the controlled device 4. The control device 1 then starts the controlled device 4 at the optimal start-up time and optimal setting value calculated by the optimization method, calculates the compressor frequency using a modern control method, and controls the controlled device 4 (here, the compressor) so that the room temperature follows the optimal stage target value. In this case, the setting value is, for example, the fan airflow rate of the fan in the indoor unit.

[0074] Furthermore, the control system 100 is not limited to an air conditioning system and can be a system other than an air conditioning system. For example, the control system 100 may be a plant system such as a manufacturing plant, a power generation plant, a petrochemical plant, or a water treatment plant. For example, in a plant system, the control device 1 may control the cutting speed of the controlled equipment 4, which is a facility or device constituting the plant system, so that the cutting amount reaches a target value (final target value) at a target time. For example, the control device 1 executes a simulation and calculates an optimal start time, an optimal setting value, and an optimal step target value for the cutting amount at the optimal step time for the controlled equipment 4 using an optimization method. The control device 1 then starts the controlled equipment 4 at the optimal start time and optimal setting value calculated using the optimization method, calculates the cutting speed using a modern control method, and controls the controlled equipment 4 (e.g., a cutting machine) so that the cutting amount follows the optimal step target value. In this case, the setting value may be, for example, the feed rate, cutting depth, cutting angle, etc. of the cutting machine.

[0075] Furthermore, for example, the control system 100 may be a robot system equipped with a transport robot (e.g., a drone). In the robot system, for example, the control device 1 may control the control target device 4, including the transport robot and related to the transport of cargo by the transport robot, so that the movement amount of the transport robot reaches a target value (final target value) at a target time. For example, the control device 1 executes a simulation and calculates an optimal start time, optimal setting value, and optimal step target value of the movement amount of the transport robot at the optimal step time using an optimization method. The control device 1 then starts the control target device 4 at the optimal start time and optimal setting value calculated using the optimization method, calculates the movement speed of the transport robot using a modern control method, and controls the control target device 4 (here, a motor, for example) so that the movement amount of the transport robot follows the optimal step target value. In this case, the setting value may be, for example, the rotation speed of a propeller provided on the transport robot or the amount of power supplied to the transport robot.

[0076] In the first embodiment, the simulator 2 is a device separate from the control device 1, but this is merely an example. For example, the simulator 2 may be provided in the control device 1.

[0077] In the first embodiment, the model creating device 3 is a device separate from the control device 1, but this is merely an example. For example, the model creating device 3 may be provided in the control device 1.

[0078] For example, both the simulator 2 and the model creating device 3 may be provided in the control device 1 .

[0079] In the first embodiment, the control device 1 and the model creation device 3 are mounted on a server, but this is merely an example. For example, the control device 1 and the model creation device 3 may be mounted on the controlled device 4. In the control device 1, for example, some of the simulation execution instruction unit 11, the simulation result collection unit 12, the optimization unit 13, the optimal control calculation unit 14, the control unit 15, and the data collection unit 16 may be provided on the server, and the rest may be provided on the controlled device 4.

[0080] In the first embodiment described above, the optimization unit 13 in the control device 1 calculates the optimal startup time of the controlled device 4 and the optimal setting value of the controlled device 4 using a known optimization algorithm. However, this is merely an example, and the optimization unit 13 may calculate the optimal startup time of the controlled device 4 and the optimal setting value of the controlled device 4 using other methods. For example, the optimization unit 13 may calculate the optimal startup time of the controlled device 4 and the optimal setting value of the controlled device 4 using reinforcement learning, which is a known machine learning method that models the relationship between the setting value and the control system and calculates the optimal setting value. However, as described above, optimization calculation is recommended as a method for calculating the optimal startup time of the controlled device 4 and the optimal setting value of the controlled device 4.

[0081] In the first embodiment described above, the optimal control calculation unit 14 in the control device 1 calculates the control variable using a known modern control algorithm based on the sensor data collected by the data collection unit 16, so as to track the optimal stage target value at the optimal stage time calculated by the optimization unit 13. However, this is merely an example, and the optimal control calculation unit 14 may calculate the control variable using, for example, other control algorithms. For example, the optimal control calculation unit 14 may calculate the control variable using a known classical control algorithm. An example of a known classical control algorithm is a PID control algorithm. Note that while a classical control algorithm enables control to track a target value, it does not take energy minimization into consideration. Therefore, for example, it is also useful in terms of energy minimization for the optimal control calculation unit 14 to calculate a stage target value that minimizes energy through optimization calculation, and then perform PID control to track that target value.

[0082] As described above, according to the first embodiment, the control device 1 is configured to include an optimization unit 13 that calculates, by optimization calculation, control data including step target values, which are step-by-step target values ​​until the controlled quantity reaches the target value, based on simulation results indicating the error between the controlled quantity and the target value and power consumption, obtained by causing the simulator unit 21, which executes a simulation that simulates the operation or behavior of the controlled device 4, to execute a simulation; an optimal control calculation unit 14 that starts the controlled device 4, collects sensor data related to the controlled quantity measured after the start of the controlled device 4, and calculates, based on the collected sensor data, a control quantity for the controlled device 4 so that the controlled quantity follows the step target value calculated by the optimization unit 13; and a control unit 15 that controls the controlled device 4 with the control quantity calculated by the optimal control calculation unit 14. Therefore, the control device 1 can perform the target control at the target time and prevent energy loss caused by the controlled quantity reaching the target value earlier than the target time. By calculating the step target value using an optimization method, the control device 1 can perform control so that the controlled quantity reaches the target value by the target time, and can prevent energy loss caused by the controlled quantity reaching the target value earlier than expected.

[0083] The control data includes the startup time of the controlled device 4 and the setting values ​​of the controlled device 4, and the optimal control calculation unit 14 starts up the controlled device at the startup time based on the control data and with the setting values. The control device 1 is capable of optimizing the startup of the controlled device 4, including the startup time and setting values ​​of the controlled device 4, which is difficult to achieve with modern control alone.

[0084] Furthermore, in the control device 1, the optimal control calculation unit 14 calculates the control amount using a system identified model in which the relationship between the control amount and power consumption and the relationship between the control amount and the controlled amount are modeled, and the system identified model is created using simulation data collected when the simulator unit 21 executes a simulation. The control device 1 can utilize data obtained in the optimization process (simulation data) for modern control. The control device 1 can contribute to shortening the time required to collect data necessary for modern control, more specifically, for creating a system identified model necessary for modern control.

[0085] Furthermore, according to the first embodiment, the control system 100 is configured to include: an optimization unit 13 that calculates, by optimization calculation, control data including stage target values, which are staged target values ​​until the controlled quantity reaches the target value, based on simulation results indicating the error between the controlled quantity and the target value and power consumption, obtained by causing a simulator unit 21 that executes a simulation that simulates the operation or behavior of the controlled device 4; an optimal control calculation unit 14 that starts the controlled device 4, collects sensor data related to the controlled quantity measured after the start of the controlled device 4, and calculates, based on the collected sensor data, a control quantity for the controlled device 4 so that the controlled quantity follows the staged target value calculated by the optimization unit 13; a control unit 15 that controls the controlled device 4 with the control quantity calculated by the optimal control calculation unit 14; and the controlled device 4. Therefore, the control system 100 can perform the target control at the target time and prevent energy loss caused by the controlled quantity reaching the target value earlier than the target time. By calculating the stage target value using an optimization method, the control system 100 can perform control so that the controlled quantity reaches the target value by the target time, and can prevent energy loss caused by the controlled quantity reaching the target value earlier than expected.

[0086] The control data includes the startup time of the controlled device 4 and the setting values ​​of the controlled device 4, and the optimal control calculation unit 14 starts up the controlled device at the startup time based on the control data and with the setting values. The control system 100 makes it possible to optimize the startup of the controlled device 4, including the startup time and setting values ​​of the controlled device 4, which is difficult to achieve with modern control alone.

[0087] Furthermore, in the control system 100, the control device 1 calculates the controlled variable using a system identified model in which the relationship between the controlled variable and power consumption and the relationship between the controlled variable and the controlled variable are modeled, and the system identified model is created using simulation data collected when the simulator unit 21 executes a simulation. The control system 100 can utilize data (simulation data) obtained during the optimization process for modern control. The control system 100 can contribute to shortening the time required to collect data necessary for modern control, more specifically, for creating a system identified model necessary for modern control.

[0088] Any of the components of the embodiments may be modified or omitted.

[0089] The control device according to the present disclosure can perform targeted control at a target time and prevent energy loss caused by the controlled variable reaching the target value earlier than the target time.

[0090] 1 Control device, 11 Simulation execution instruction unit, 12 Simulation result collection unit, 13 Optimization unit, 14 Optimal control calculation unit, 15 Control unit, 16 Data collection unit, 2 Simulator, 21 Simulator unit, 22 Data storage unit, 3 Model creation device, 31 Model creation unit, 32 Model storage unit, 4 Control target equipment, 100 Control system, 1001 Processing circuit, 1002 Input interface device, 1003 Output interface device, 1004 Processor, 1005 Memory.

Claims

1. A control device that controls a controlled device to be controlled so that a controlled quantity, which is a measurement value aiming for a target value, reaches the target value at a target time, comprising: an optimization unit that calculates, by optimization calculation, control data including stage target values, which are staged target values ​​until the controlled quantity reaches the target value, based on a simulation result indicating an error between the controlled quantity and the target value and power consumption, obtained by having a simulator unit that executes a simulation that simulates the operation or behavior of the controlled device execute the simulation; an optimal control calculation unit that starts up the controlled device, collects sensor data related to the controlled quantity measured after the control target device is started up, and calculates a control quantity for the controlled device based on the collected sensor data so that the controlled quantity follows the stage target value calculated by the optimization unit; and a control unit that controls the controlled device with the control quantity calculated by the optimal control calculation unit.

2. The control device according to claim 1, characterized in that the control data includes a start-up time of the controlled device and a setting value of the controlled device, and the optimal control calculation unit starts up the controlled device at the start-up time based on the control data and with the setting value.

3. A control device according to claim 1 or 2, characterized in that the optimal control calculation unit calculates the controlled variable using modern control.

4. A control device as claimed in any one of claims 1 to 3, characterized in that the optimal control calculation unit calculates the control quantity using a system identified model in which the relationship between the control quantity and the power consumption and the relationship between the control quantity and the controlled quantity are modeled, and the system identified model is created using simulation data collected when the simulator unit executes the simulation.

5. A control device as claimed in any one of claims 1 to 4, characterized in that the optimization unit is capable of selecting, in the optimization calculation, multi-objective optimization that can select comfort in which the measured value and the target value match, energy saving, or a trade-off in power consumption.

6. The control device according to any one of claims 1 to 5, further comprising: a simulator section for executing the simulation and outputting a result of the simulation.

7. A control system comprising: a control device according to any one of claims 1 to 6; and the controlled device.

8. A control system comprising: a control device according to any one of claims 1 to 5; a simulator section which executes the simulation and outputs the simulation results; and a device to be controlled.

9. The control system according to claim 7 or 8, which is an air conditioning system.

10. A control method for controlling a controlled device to be controlled so that a controlled quantity, which is a measurement value aiming for a target value, reaches the target value at a target time, comprising the steps of: an optimization unit calculating, by optimization calculation, control data including stage target values, which are staged target values ​​until the controlled quantity reaches the target value, based on a simulation result indicating an error between the controlled quantity and the target value and power consumption, obtained by having a simulator unit that executes a simulation to simulate the operation or behavior of the controlled device execute the simulation; an optimal control calculation unit starting up the controlled device, collecting sensor data related to the controlled quantity measured after the start-up of the controlled device, and calculating, based on the collected sensor data, a control quantity for the controlled device so that the controlled quantity follows the stage target value calculated by the optimization unit; and a control unit controlling the controlled device with the control quantity calculated by the optimal control calculation unit.

Images