Environmental control system and environmental control method

Abstract

Even when controlling the spatial environment using environmental prediction, the control can be performed more reliably in real time. [Solution] The environmental control system 10 is a system for controlling the environment of a space, and comprises a distribution acquisition unit 11 that acquires control distribution information indicating the probability distribution of control content and target distribution information indicating the probability distribution of the environment of the space that is the target of control, a prediction determination unit 12 that determines multiple control content to be used to predict the environment of the space according to the acquired control distribution information, a prediction unit 13 that predicts the environment of the space after control for each of the multiple determined control content when control is performed with that content, and a determination unit 14 that compares the predicted environment of the space after control for each of the multiple control content with the probability distribution of the environment indicated by the acquired target distribution information, and determines the control content for the space in accordance with the comparison.

Description

[Technical Field]

[0001] The present invention relates to an environmental control system and an environmental control method for controlling the environment of a space. [Background technology]

[0002] BACKGROUND ART Conventionally, it has been proposed to predict the environment (state), such as the future temperature, of a space to be air-conditioned, and to control the air conditioning in accordance with the prediction (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-37177 Summary of the Invention [Problem to be solved by the invention]

[0004] One way to control air conditioning using a prediction of the environment of a space to be air-conditioned is to optimize the control in accordance with the prediction. Control optimization is performed, for example, as follows: The environment of the space after control (e.g., space temperature) when air conditioning is performed using specific control content (e.g., operation variables such as supply air temperature and air volume) is predicted. It is determined whether the predicted environment is in a predetermined target state. If it is determined that the predicted environment is in the target state, the control content used for the prediction is determined to be the optimal control content. On the other hand, if the predicted environment is not in the target state, the control content is changed according to the control content used for the prediction, and the changed control content is used to predict the environment of the space after control. By repeating prediction while changing the control content in this way, a search is made for control content that can bring the environment of the space to the target state.

[0005] By performing this type of optimization, the space to be air-conditioned can be brought to the target state more quickly than with feedback control such as PID (Proportional Integral Derivative) control. Also, by performing this type of optimization, air conditioning to bring the space to the target state can be performed even in locations where sensors that detect the environment (for example, temperature sensors) are not installed in the space.

[0006] However, the above optimization requires repeated predictions of the spatial environment, which places a heavy burden on the calculations required to determine the optimal control. As a result, the calculation load required to determine the control cannot keep up with the actual phenomena, making real-time control difficult. In particular, when a high-load calculation model such as CFD (computational fluid dynamics) is used to make an appropriate prediction, real-time control is difficult.

[0007] The present invention has been made in consideration of the above, and aims to provide an environmental control system and an environmental control method that can more reliably control the environment of a space in real time, even when controlling the environment using environmental prediction. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, the environmental control system of the present invention is an environmental control system that controls the environment of a space, and comprises: a distribution acquisition means that acquires control distribution information that indicates a probability distribution of control content and target distribution information that indicates a probability distribution of the environment of the space that is the target of control; a prediction determination means that determines multiple control content to be used to predict the environment of the space in accordance with the control distribution information acquired by the distribution acquisition means; a prediction means that predicts the environment of the space after control for each of the multiple control content determined by the prediction determination means when control is performed with that content; and a determination means that compares the environment of the space after control for each of the multiple control content predicted by the prediction means with the probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition means, and determines the control content for the space in accordance with the comparison.

[0009] In the environmental control system according to the present invention, multiple control details to be used in predicting the environment of a space are determined according to control distribution information, and the post-control environment of the space when control is performed using each of the determined multiple control details is predicted. The predicted post-control environment of the space for each of the multiple control details is compared with the probability distribution of the environment indicated by the target distribution information, and the control details for the space are determined based on the comparison. The prediction of the post-control environment of the space for each of the multiple control details can be made without assuming other control details. Therefore, unlike the case where predictions are made sequentially, such as in the control optimization described above, each prediction can be made in parallel, allowing for faster predictions. As a result, the control details can be determined quickly. Therefore, according to the environmental control system according to the present invention, even when controlling the environment of a space using environmental predictions, control can be performed more reliably in real time.

[0010] The control contents may include a plurality of items. With this configuration, it is possible to more reliably control a plurality of items in real time.

[0011] The environment of the space to be controlled may be at least one of temperature, humidity, wind speed, and particle concentration. With this configuration, at least one of the temperature, humidity, wind speed, and particle concentration of the space can be controlled more reliably in real time.

[0012] The prediction means may acquire information indicating the measurement results of the environment of the space, and predict the environment of the space after control based on the acquired information. With this configuration, appropriate control can be performed according to the measurement results of the environment of the space.

[0013] The determining means may determine whether or not to perform the control with the determined control content based on the comparison. With this configuration, the control with the determined content can be performed only when the control with the determined content is appropriate.

[0014] Incidentally, the present invention can be described not only as an invention of an environmental control system as described above, but also as an invention of an environmental control method as described below. These are essentially the same inventions, with similar actions and effects, but in different categories.

[0015] That is, the environmental control method of the present invention is an environmental control method that is an operating method of an environmental control system that controls the environment of a space, and includes: a distribution acquisition step that acquires control distribution information that indicates the probability distribution of control content and target distribution information that indicates the probability distribution of the environment of the space that is the target of the control; a prediction determination step that determines multiple control content to be used for predicting the environment of the space in accordance with the control distribution information acquired in the distribution acquisition step; a prediction step that predicts the environment of the space after control for each of the multiple control content determined in the prediction determination step when control is performed with that content; and a determination step that compares the environment of the space after control for each of the multiple control content predicted in the prediction step with the probability distribution of the environment indicated by the target distribution information acquired in the distribution acquisition step, and determines the control content for the space in accordance with the comparison. [Effects of the Invention]

[0016] According to the present invention, even when controlling the environment of a space using environmental prediction, it is possible to perform the control more reliably in real time. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a diagram showing a configuration of an environmental control system according to an embodiment of the present invention. [Figure 2] 10A to 10C are diagrams showing examples of control distribution information and target distribution information, examples of predicted blown-out temperature and blown-out air volume, and examples of predicted space temperature. [Figure 3] FIG. 10 is a diagram showing predicted values ​​of space temperatures at each position in a space to be controlled. [Figure 4] 3 is a flowchart illustrating an environmental control method, which is a process executed in the environmental control system according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0018] Hereinafter, an embodiment of an environmental control system and an environmental control method according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same elements are given the same reference numerals and redundant description will be omitted.

[0019] FIG. 1 shows an environmental control system 10 according to this embodiment. The environmental control system 10 is a system (device) that controls the environment (state) of a space. The environment to be controlled in the space is, for example, the thermal environment of the space, such as the space temperature. In the following description, as an example, the environmental control system 10 will be described as controlling the space temperature. Note that the environment to be controlled may be something other than the space temperature. For example, the environment to be controlled may be humidity, wind speed, or particle concentration. Furthermore, the environment to be controlled may be more than one of the above. If the environment to be controlled is something other than the space temperature, then in the following description, the space temperature may be read as something other than the space temperature.

[0020] The space to be controlled (the space in which environmental control is performed) is, for example, a room in a building. When the space to be controlled has temporal and spatial heat distribution, the control according to this embodiment is particularly effective. For example, the space to be controlled is a space (room) in which servers in a data center are installed. In this space, there is uneven heat distribution due to the large number of server racks. In this space, stable temperature control is required 24 hours a day, 365 days a year. In this space, energy-saving operation is performed through highly accurate temperature control.

[0021] Alternatively, the space to be controlled may be a space where factory equipment is installed. In this space, manufacturing equipment that generates heat is irregularly arranged. In this space, it is necessary to ensure the comfort of workers. In this space, temperature control is required to maintain product quality. Alternatively, the space to be controlled may be an office space. In this space, there is a fluctuating heat distribution due to the flow of people. In this space, it is necessary to eliminate the temperature difference between the window area and the interior. In this space, it is necessary to achieve both energy conservation and comfort.

[0022] Specifically, the environmental control system 10 is a computer including hardware such as a CPU (Central Processing Unit) and memory. Each function of the environmental control system 10, which will be described later, is realized by these components operating through programs or the like. The environmental control system 10 has a configuration capable of parallel processing. The environmental control system 10 may be realized by a single computer, or may be realized by a computer system configured by connecting multiple computers to each other via a network. The environmental control system 10 may have a communication function for acquiring information necessary for controlling the environment and for controlling the environment, as described below.

[0023] As shown in FIG. 1, an air conditioner 20 is positioned and fixedly installed in a space to be controlled. The air conditioner 20 performs air conditioning by blowing out conditioned air into the space to be controlled. The air conditioner 20 may be a conventional air conditioner such as an AHU (air handling unit). The space to be controlled may be installed with multiple air conditioners 20, each of which can be controlled individually. The air conditioners 20 are installed in positions where they can perform appropriate air conditioning in the space to be controlled. When multiple air conditioners 20 are installed, the air conditioners 20 are installed at positions spaced at regular intervals, for example.

[0024] The environmental control system 10 controls the space temperature by controlling the operation of the air conditioner 20. The air conditioner 20 is capable of setting and controlling the temperature of the air blown out (discharge temperature, supply air temperature) and air volume (discharge air volume). The air conditioner 20 is connected to the environmental control system 10 by wire or wirelessly, and is capable of being controlled by the environmental control system 10. Information may be exchanged between the environmental control system 10 and the air conditioner 20 via a control interface such as BACnet (Building Automation and Control Networking protocol) or Modbus.

[0025] The environmental control system 10 controls the space temperature by setting and controlling the air outlet temperature and air volume (operation amount of the air conditioner 20) of the air conditioner 20. Note that in the following example, an example will be described in which the air outlet temperature and air volume of the air conditioner 20 are used as the control items of the environmental control system 10, but the control items are not necessarily limited to the above. For example, the control items may include the wind direction of the air blown out from the air conditioner 20. Note that a device other than the air conditioner 20 (a device other than one that blows out conditioned air to perform air conditioning) may be used as a device for controlling the environment of the space.

[0026] As shown in FIG. 1, a sensor 30 is positioned and fixedly installed in a space to be controlled. The sensor 30 is a device that measures (detects) the environment of the controlled object at the position where the sensor 30 is installed. If the environment of the controlled object is space temperature, as in this embodiment, the sensor 30 measures the space temperature. The sensor 30 may be a conventional sensor such as a conventional temperature sensor. A plurality of sensors 30 may be installed in the space to be controlled. The position where the sensor 30 is installed is a position where appropriate control can be performed by measuring the environment (space temperature). When a plurality of sensors 30 are installed, the sensors 30 are installed at positions, for example, at regular intervals.

[0027] The environmental control system 10 uses the space temperature measured by the sensor 30 to control the space temperature. The sensor 30 is connected to the environmental control system 10 by wire or wirelessly and is capable of transmitting information to the environmental control system 10. The sensor 30 transmits information indicating the measured space temperature to the environmental control system 10.

[0028] If the environment to be controlled is humidity, wind speed, or particle concentration, the sensor 30 measures the humidity, wind speed, or particle concentration. Alternatively, a sensor that measures an environment other than the environment to be controlled may be provided in the space to be controlled, and the environmental control system 10 may use information obtained by measurement by the sensor to control the environment.

[0029] Next, the functions of the environmental control system 10 according to this embodiment will be described. As shown in Fig. 1, the environmental control system 10 includes a distribution acquisition unit 11, a prediction determination unit 12, a prediction unit 13, and a determination unit 14.

[0030] The control of space temperature by the environmental control system 10 is performed by repeatedly determining the values ​​of the air conditioner 20's blowing temperature and blowing air volume at regular time intervals (for example, several minutes to several tens of minutes) and setting them for the air conditioner 20. The function of the environmental control system 10 described below is a function for determining and controlling the values ​​of the air conditioner 20's blowing temperature and blowing air volume for each successive time step at the above-mentioned time intervals.

[0031] The distribution acquisition unit 11 is a distribution acquisition means for acquiring control distribution information indicating a probability distribution of the control content and target distribution information indicating a probability distribution of the environment of the space that is the target of the control. The control content may have multiple items.

[0032] The control distribution information is information indicating the probability distribution (prior distribution) of the control content. The control content is, for example, the values ​​of the air outlet temperature and air volume of the air conditioner 20 to be determined. The probability distribution of the control content is, for example, the distribution of the probability that each of these values ​​will be selected as the value to be used in the prediction described below. The control distribution information is information for each control item (operation variable), i.e., for each air outlet temperature and air volume. The control distribution information may be information for each time step (information that differs for each time step).

[0033] An example of the control distribution information is shown in Figure 2. The graph of the blowing temperature and blowing air volume shown in Figure 2 corresponds to the control distribution information. The coordinate axes of the graph are the axes of the blowing temperature and blowing air volume values, respectively. The values ​​of the graph for each value of the blowing temperature and blowing air volume are the above-mentioned probability values.

[0034] The distribution acquisition unit 11 acquires control distribution information by generating it for each time step, for example, as follows. The control distribution information may correspond to the value set for the air conditioner 20 in the previous time step. The distribution acquisition unit 11 acquires the value determined by the determination unit 14 and set for the air conditioner 20 in the previous time step. The distribution acquisition unit 11 sets a predetermined and stored type of probability distribution (e.g., Gaussian distribution) with the acquired value as the average value as the control distribution information. If there is no value determined in the previous time step, i.e., in the case of the first time step, it is sufficient to use a predetermined and stored value as the average value.

[0035] When the preset probability distribution is a Gaussian distribution, the distribution acquisition unit 11 also stores in advance variations such as standard deviation (variance). For example, the standard deviation is set according to the range in which the values ​​of the air conditioner 20's blowout temperature and blowout air volume can be changed from the previous time step, in terms of the performance of the air conditioner 20. Specifically, the standard deviation is set so that the probability distribution that is the control distribution information falls roughly within the range in which the values ​​of the air conditioner 20's blowout temperature and blowout air volume can be changed from the previous time step (settable range).

[0036] Setting of the control distribution information in this manner takes into consideration changes in the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 since the previous time step. If it is not necessary to take this into consideration, the control distribution information does not necessarily have to be as described above, and it will suffice if it is an appropriate probability distribution of the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 to be determined.

[0037] The probability distribution associated with the control distribution information may differ for each control item, i.e., for each blowout temperature and blowout air volume. For example, the variance of the probability distribution may be set according to the preset priority or tolerance of the control item. Specifically, the variance of the probability distribution may be increased for items with high priority and decreased for items with low priority. Alternatively, the variance of the probability distribution may be decreased for items with high priority and increased for items with low priority.

[0038] When multiple air conditioners 20 are installed in the space to be controlled, the distribution acquisition unit 11 acquires control distribution information for each air conditioner 20. Note that the control distribution information does not necessarily have to be the one described above, as long as it indicates a probability distribution of the control content. The distribution acquisition unit 11 may also acquire control distribution information by a method other than generation. For example, the distribution acquisition unit 11 may acquire control distribution information by reading it out from control distribution information that has been set and stored in advance by a user. In this case, the control distribution information may be information set for each time step, or may be uniform information regardless of the time step.

[0039] The target distribution information is information indicating the probability distribution of the environment (target state) of the space that is the target of control. The probability distribution of the environment of the space that is the target of control is, for example, the distribution of the probability that each value of the space temperature will be the target of control. The higher the probability, the more desirable the value of the space temperature is. The target distribution information may be information for each time step (information that differs for each time step). The target distribution information may also be information for each position of the space to be controlled (information that differs for each position of the space to be controlled). In this case, the position is, for example, each position obtained by dividing the space to be controlled into a grid (mesh).

[0040] Figure 2 shows an example of target distribution information for one position. The spatial temperature graph shown in Figure 2 corresponds to the target distribution information. The coordinate axis of the graph is the axis of spatial temperature values. The values ​​of the graph for each spatial temperature value are the above-mentioned probability values.

[0041] The distribution acquisition unit 11 acquires the target distribution information by generating it for each time step, for example, as follows: The distribution acquisition unit 11 acquires a set value of the space temperature (e.g., 25°C). The set value of the space temperature is set by the user for the environmental control system 10. The set value of the space temperature may be a value for each position in the space to be controlled. The distribution acquisition unit 11 sets the acquired value as the average value and sets a predetermined and stored type of probability distribution (e.g., Gaussian distribution) as the target distribution information.

[0042] If the preset probability distribution is a Gaussian distribution, the distribution acquisition unit 11 also stores in advance variations such as standard deviation (variance). For example, the standard deviation is set depending on how much error is allowed from the set value of the space temperature. Specifically, the standard deviation (for example, 1°C) is set so that the probability distribution that is the target distribution information falls roughly within the range of allowable error.

[0043] Note that the target distribution information does not necessarily have to be the above, as long as it indicates the probability distribution of the environment of the space that is the target of control. Furthermore, the distribution acquisition unit 11 may acquire the target distribution information by a method other than generation. For example, the distribution acquisition unit 11 may acquire target distribution information by reading out target distribution information that has been set and stored in advance by a user. In this case, the target distribution information may be information set for each time step, or may be uniform information regardless of the time step. The target distribution information set by the user may be, for example, based on a space usage schedule or external information (e.g., weather forecast or electricity price).

[0044] The distribution acquisition unit 11 outputs the acquired control distribution information to the prediction determination unit 12. The distribution acquisition unit 11 outputs the acquired target distribution information to the determination unit .

[0045] The prediction determining unit 12 is a prediction determining means that determines the contents of a plurality of controls to be used for predicting the environment of the space in accordance with the control distribution information acquired by the distribution acquiring unit 11.

[0046] The plurality of control contents determined by the prediction determining unit 12 are used to predict the environment of the space after the control, which will be described later. The plurality of control contents are control patterns for each prediction. The prediction determining unit 12 determines control contents related to a preset number of control patterns (the number corresponding to the number of predictions).

[0047] The forecast-use determination unit 12 makes a determination, for example, as follows. The forecast-use determination unit 12 stores determination rules in advance and makes a determination in accordance with the rules. The forecast-use determination unit 12 inputs control distribution information from the distribution acquisition unit 11. The forecast-use determination unit 12 determines the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 for each control pattern with a probability according to the probability distribution indicated in the control distribution information. The values ​​of multiple points on the coordinate axes of the blowing temperature and blowing air volume shown in FIG. 2 are examples of the values ​​of the blowing temperature and blowing air volume determined for each control pattern. When the space to be controlled includes multiple air conditioners 20, the forecast-use determination unit 12 determines the values ​​of the blowing temperature and blowing air volume for each air conditioner 20.

[0048] The prediction determination unit 12 generates control patterns by determining the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 for the above-mentioned preset number (the number of control patterns). The prediction determination unit 12 outputs to the prediction unit 13 information indicating the determined multiple control contents, i.e., the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 for each of the multiple control patterns.

[0049] The prediction unit 13 is a prediction means for predicting the environment of the space after control when control is performed with each of the multiple control contents determined by the prediction determination unit 12. The prediction unit 13 may obtain information indicating the measurement results of the environment of the space, and predict the environment of the space after control based on the obtained information.

[0050] The prediction of the post-control space environment by the prediction unit 13 is used to determine the content of control. The prediction unit 13 makes a prediction, for example, as follows. The prediction unit 13 stores prediction rules in advance and makes a prediction in accordance with the rules. The prediction unit 13 inputs information indicating the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 for each of a plurality of control patterns from the prediction determination unit 12. The prediction unit 13 predicts the post-control space environment, assuming that the control indicated by the input information is performed on the air conditioner 20 and the air conditioner 20 operates, for each control pattern. The prediction unit 13 predicts the space temperature of the space to be controlled in the next time step as the post-control space environment.

[0051] The prediction unit 13 predicts the space temperature at a predetermined position in the space to be controlled. The prediction unit 13 also predicts the space temperature at multiple positions in the space to be controlled. The multiple positions in this case are positions corresponding to the target distribution information described above. The positions in this case are, for example, each position obtained by dividing the space to be controlled into a grid (mesh). The values ​​of multiple points on the space temperature coordinate axis shown in Figure 2 are examples of space temperature values ​​predicted for each control pattern. Figure 3 also shows predicted space temperature values ​​for each position in the space to be controlled. In Figure 3, the color intensity represents the temperature value. The "AHU" part in the figure indicates the position where the air conditioner 20 is installed, and the "T" part indicates the position where the sensor 30 is installed.

[0052] The prediction unit 13 uses the space temperature of the space to be controlled at the current time step for the prediction. Therefore, the prediction unit 13 predicts the space temperature of the space to be controlled at the current time step from the space temperature of the space to be controlled at the previous time step. For this prediction, the prediction unit 13 predicts the space temperature at the current time step, assuming that the control determined by the determination unit 14 was performed on the air conditioner 20 at the previous time step and the air conditioner 20 operated. This prediction may be performed in the same manner as the prediction of the space temperature of the space to be controlled at the next time step. The predicted space temperature at the current time step is used to predict the space temperature at that time in the next step.

[0053] In addition, if there is no space temperature of the space to be controlled for the previous time step, i.e., if the current time step is the first time step, the prediction unit 13 may use a preset value (e.g., a statistical value of past measurement values) as the predicted value.

[0054] The prediction by the prediction unit 13 may be performed by a conventional method. For example, the prediction may be performed by a numerical model simulator that executes a numerical model that reproduces the thermal and airflow environment of the space to be controlled. The prediction unit 13 may also use information related to the space to be controlled and information related to the air conditioner 20 for the prediction. This information is, for example, information on the shape and material of the space, the specifications of the air conditioner 20, and boundary conditions (for example, information related to weather such as outside temperature and solar radiation). When the prediction unit 13 uses this information for the prediction, it acquires the information before the prediction. The information may be acquired by an existing method, such as accepting an input operation from a user or receiving information from the outside (for example, a weather forecast).

[0055] The prediction unit 13 may use, as the space temperature of the space to be controlled at the current time step, the space temperature based on the measurement by the sensor 30. In this case, the prediction unit 13 receives and acquires information indicating the space temperature transmitted from the sensor 30.

[0056] As described above, when the positions of the prediction target are each position obtained by dividing the space of the control target into a grid, sensors 30 are not usually installed at all of these positions. Therefore, the prediction unit 13 predicts the spatial temperature of the space of the control target at each position at the current time step by data assimilation calculation based on the prediction result of the spatial temperature at the current time step and the measurement result (actual measured value) of the spatial temperature at the position where the sensor 30 is installed, which is indicated by the information received from the sensor 30. The data assimilation calculation may be performed in the same manner as conventional (for example, calculation using an ensemble Kalman filter or a particle filter).

[0057] By performing data assimilation calculations using the measurement results of the sensor 30, the predicted spatial temperature can be corrected taking into account prediction errors and measurement errors of the numerical model, thereby obtaining a more likely spatial temperature at the current time step. Note that if sensors 30 are installed at all positions of the prediction target, the prediction unit 13 may use the measurement results of the spatial temperature at the positions where the sensor 30 is installed as the spatial temperature at the current time step, without predicting the spatial temperature at the current time step.

[0058] The above-described multiple control patterns are independent of each other. Therefore, the prediction unit 13 can predict the space temperature independently for each control pattern. The prediction unit 13 predicts the space temperature in parallel for each control pattern. By predicting the space temperature in parallel, the time required to determine the blowout temperature and blowout air volume of the air conditioner 20 can be shortened. This ensures real-time control of the space temperature. Furthermore, the prediction unit 13 may perform predictions using a high-load calculation model such as CFD, which could not be used in the case of the above-described conventional repeated predictions due to the calculation load.

[0059] The prediction unit 13 outputs to the determination unit 14 information indicating the space temperature of the space to be controlled in the next time step, which is the prediction result for each of the plurality of control patterns.

[0060] The determination unit 14 is a determination means that compares the environment of the space after control for each of the plurality of control contents predicted by the prediction unit 13 with the probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition unit 11, and determines the control contents for the space in accordance with the comparison. The determination unit 14 may determine whether or not to perform control with the control contents to be determined in accordance with the comparison.

[0061] The determination unit 14 determines the content of control for the space to be controlled, for example, as follows. The determination unit 14 stores decision rules in advance and makes decisions in accordance with the rules. The determination unit 14 inputs target distribution information from the distribution acquisition unit 11. The determination unit 14 inputs information indicating prediction results for each of a plurality of control patterns from the prediction unit 13. The determination unit 14 determines the content of control for the space to be controlled from the target distribution information and the prediction results so that the target related to the target distribution information is achieved. The determination unit 14 determines, for example, the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 at the current time step as the content of control for the space to be controlled.

[0062] The determination unit 14 compares the prediction results for each of the multiple control patterns (values ​​of multiple points on the coordinate axis of space temperature shown in FIG. 2) with the probability distribution of space temperature indicated by the target distribution information (the graph of space temperature shown in FIG. 2), and determines the values ​​of the blowing temperature and blowing air volume of the air conditioner 20 at the current time step. This determination is made, for example, so that the determined values ​​of the blowing temperature and blowing air volume of the air conditioner 20 are values ​​that are considered to be most likely to achieve the probability distribution of space temperature indicated by the target distribution information.

[0063] For example, this determination is made by data assimilation calculation. The data assimilation calculation here estimates the control content to realize the target state by assimilating the target state indicated by target distribution information, rather than observation data, into the prediction. The data assimilation calculation is, for example, a method using likelihood calculation and resampling in a particle filter, or a method using an ensemble Kalman filter using an extended state vector. The assimilation calculation used for this determination is, for example, that shown in Morishita, Y., Murakami, S., Yokoyama, M., Ueno, G., 2023. Data assimilation and control system for adaptive model predictive control. Journal of Computational Science 72, 102079 (Non-Patent Document 1).

[0064] Specifically, the determination unit 14 calculates the likelihood of realizing the probability distribution of the spatial temperature indicated by the target distribution information for the predicted results for each control pattern, and calculates the probability distribution of the blowing temperature and blowing air volume for determining the blowing temperature and blowing air volume of the air conditioner 20 at the current time step based on the likelihood for each control pattern.

[0065] The determination unit 14 determines the blowing temperature and blowing air volume of the air conditioner 20 for the current time step based on the calculated probability distribution. For example, the determination unit 14 determines the blowing temperature and blowing air volume of the air conditioner 20 with the highest probability in the probability distribution as the blowing temperature and blowing air volume of the air conditioner 20 for the current time step. The determination may also use an expected value or a mode of the probability distribution. The determination may also be made by performing an optimization calculation taking into account the constraints of the air conditioner 20 (configurable range, response speed) and an evaluation index for energy saving.

[0066] Note that if the target distribution information is information for each of a plurality of positions (if there are multiple graphs of space temperature as shown in FIG. 2), the target distribution information for each of the plurality of positions is used collectively in the data assimilation calculation. Also, if there are multiple air conditioners 20 for which the blowing temperature and blowing air volume are to be determined (if there are multiple graphs of blowing temperature and blowing air volume as shown in FIG. 2), the blowing temperature and blowing air volume for each of the plurality of air conditioners 20 are determined collectively by the data assimilation calculation. For example, the determination unit 14 treats the combinations of the blowing temperatures and blowing air volumes of all of the plurality of air conditioners 20 as a single vector and determines the blowing temperature and blowing air volume that are likely to achieve the probability distribution of the target state for the entire space.

[0067] The determination unit 14 controls the air conditioner 20 to operate at the determined values ​​of the discharge temperature and discharge air volume by transmitting information indicating the determined values ​​of the discharge temperature and discharge air volume to the air conditioner 20 and setting the same. The air conditioner 20, which is controlled by the environmental control system 10, operates in accordance with the control. The determined values ​​of the discharge temperature and discharge air volume as described above are used to determine the values ​​of the discharge temperature and discharge air volume in the next time step. Note that the determination of the control content by the determination unit 14 may be performed by a method other than the above, as long as it compares the post-control spatial environment for each of the multiple control content predicted by the prediction unit 13 with the probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition unit 11 and is performed in accordance with the comparison.

[0068] Furthermore, the decision unit 14 may determine whether or not to control the air conditioner 20 with the values ​​of the blowing temperature and blowing air volume determined as described above. For example, when it is considered that the determined values ​​of the blowing temperature and blowing air volume can (highly likely to) realize the probability distribution of the space temperature indicated by the target distribution information, the decision unit 14 determines to control the air conditioner 20 with the determined values ​​of the blowing temperature and blowing air volume, and when it is considered that the determined values ​​of the blowing temperature and blowing air volume cannot (highly likely to) realize the probability distribution of the space temperature indicated by the target distribution information, the decision unit 14 determines not to control the air conditioner 20 with the determined values ​​of the blowing temperature and blowing air volume.

[0069] For example, the determination unit 14 predicts the post-control space environment when control is performed using the determined values ​​of the blow-out temperature and blow-out air volume. Specifically, the determination unit 14 predicts the space temperature of the space to be controlled at the next time step as the post-control space environment. This prediction may be performed in the same manner as the prediction by the prediction unit 13 described above. Alternatively, the prediction by the prediction unit 13 may be performed at this timing, and the determination unit 14 may use the prediction by the prediction unit 13.

[0070] The decision unit 14 compares the predicted space temperature with the probability distribution of the target state indicated by the target distribution information. The decision unit 14 determines whether they are similar. If they are similar, the decision unit 14 determines that the determined values ​​of the blowing temperature and blowing air volume are likely to be able to realize the probability distribution of the space temperature indicated by the target distribution information (high probability), and determines to control the air conditioner 20 at the determined values ​​of the blowing temperature and blowing air volume. If they are not similar, the decision unit 14 determines that the determined values ​​of the blowing temperature and blowing air volume are likely to be unable to realize the probability distribution of the space temperature indicated by the target distribution information (high probability), and determines not to control the air conditioner 20 at the determined values ​​of the blowing temperature and blowing air volume.

[0071] For example, the determination unit 14 calculates the difference between the predicted space temperature and the average value (set value) of the target state as an index value indicating the similarity. The determination unit 14 compares the calculated difference with a preset threshold, and determines that the two are similar if the difference is equal to or smaller than the threshold, and determines that the two are not similar if the difference exceeds the threshold. The determination of the similarity may be based on a probabilistic likelihood.

[0072] If it is determined that the air conditioner 20 should be controlled at the determined values ​​of the discharge temperature and discharge air volume, the decision unit 14 controls the air conditioner 20 so that the air conditioner 20 operates at the values ​​of the discharge temperature and discharge air volume determined as described above. If it is determined that the air conditioner 20 should not be controlled at the determined values ​​of the discharge temperature and discharge air volume, the decision unit 14 determines the values ​​of the discharge temperature and discharge air volume by a method other than the above, and controls the air conditioner 20 so that the air conditioner 20 operates at the determined values ​​of the discharge temperature and discharge air volume. For example, in this case, the decision unit 14 controls the air conditioner 20 by normal feedback control.

[0073] The determination of whether or not to perform control with the control content determined by the determination unit 14 may be made by a method other than the above, as long as the post-control spatial environment for each of the plurality of control contents predicted by the prediction unit 13 is compared with the probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition unit 11, and the determination is made based on the comparison. The above are the functions of the environmental control system 10 according to this embodiment.

[0074] Next, an environmental control method, which is a process executed by the environmental control system 10 according to this embodiment (an operating method performed by the environmental control system 10), will be described using the flowchart of FIG. 4. This process is repeatedly performed for each time step. In this process, the distribution acquisition unit 11 acquires control distribution information indicating a probability distribution of control contents and target distribution information indicating a probability distribution of the environment of a space that is a target of control (S01, distribution acquisition step). Next, the prediction determination unit 12 determines a plurality of control contents to be used for predicting the environment of the space according to the control distribution information acquired by the distribution acquisition unit (S02, prediction determination step).

[0075] Next, the prediction unit 13 predicts the post-control environment of the space when control is performed using each of the multiple control contents determined by the prediction determination unit 12 (S03, prediction step). Next, the determination unit 14 compares the post-control environment of the space for each of the multiple control contents predicted by the prediction unit 13 with the probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition unit, and determines the control content for the space in accordance with the comparison (S04, determination step). Next, the determination unit 14 implements control using the determined control content, and the air conditioner 20 operates using the control (S05). This completes the environmental control method according to this embodiment.

[0076] In this embodiment, multiple control contents (e.g., the blowout temperature and blowout air volume of the air conditioner 20) used to predict the environment of the space are determined according to the control distribution information, and for each of the multiple determined control contents, the environment of the space after control (e.g., the space temperature) when control is performed with that content is predicted. The predicted environment of the space after control for each of the multiple control contents is compared with the probability distribution of the environment indicated by the target distribution information, and the control content for the space is determined based on the comparison.

[0077] As described above, the prediction of the spatial environment after control for each of the above multiple control contents can be performed without assuming the content of other control. Therefore, unlike the case where predictions are performed sequentially, such as in the case of control optimization described above, each prediction can be performed in parallel, allowing for faster predictions. As a result, the content of control can be determined quickly. Therefore, according to this embodiment, even when controlling the spatial environment using environmental prediction, control can be performed more reliably in real time.

[0078] Furthermore, since predictions can be made quickly, predictions can be made using a high-load calculation model, enabling highly accurate environmental control. Furthermore, for example, in controlling the environment, such as the temperature of a space, a certain degree of uncertainty may be acceptable. In response to this, in this embodiment, the target state in the space can be set as a probability distribution by using target distribution information. This allows for robust environmental control with a certain margin, taking into account the allowable error.

[0079] As in this embodiment, the control content may include multiple items (for example, the blowing temperature and blowing air volume of the air conditioner 20). This configuration makes it possible to more reliably control multiple items in real time. However, the control content may also be a single item (for example, either the blowing temperature or blowing air volume of the air conditioner 20).

[0080] As in the present embodiment, the spatial environment to be controlled may be at least one of temperature, humidity, wind speed, and particle concentration. This configuration allows for more reliable real-time control of at least one of the temperature, humidity, wind speed, and particle concentration of the space. However, the spatial environment to be controlled may be other than those described above.

[0081] As in this embodiment, the prediction unit 13 may acquire information indicating the measurement results of the spatial environment (for example, information obtained by measurement by the sensor 30) and predict the spatial environment after control based on the acquired information. Specifically, the prediction may be performed by data assimilation calculation as described above. With this configuration, appropriate control can be performed according to the measurement results of the spatial environment. However, the information indicating the measurement results of the spatial environment does not necessarily need to be used for the prediction.

[0082] As in the present embodiment, the determination unit 14 may determine whether or not to perform control with the determined control content based on the above-described comparison. With this configuration, control with the determined content can be performed only when the control with the determined content is appropriate. However, this determination does not necessarily have to be made.

[0083] The environmental control system and the environmental control method of the present disclosure have the following configuration. [1] An environmental control system that controls the environment of a space, a distribution acquisition means for acquiring control distribution information indicating a probability distribution of the content of control and target distribution information indicating a probability distribution of the environment of a space that is a target of control; a prediction determining means for determining a plurality of control contents to be used for predicting the environment of the space in accordance with the control distribution information acquired by the distribution acquiring means; a prediction means for predicting a post-control environment of a space when control is performed using each of the plurality of control contents determined by the prediction determination means; a determination means for comparing the environment of the space after control for each of the plurality of control contents predicted by the prediction means with a probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition means, and determining the control contents for the space in accordance with the comparison; An environmental control system comprising: [2] The environmental control system according to [1], wherein the control contents include a plurality of items. [3] The environmental control system according to [1] or [2], wherein the environment of the space to be controlled is at least one of temperature, humidity, wind speed, and particle concentration. [4] The environmental control system according to any one of [1] to [3], wherein the prediction means acquires information indicating the measurement results of the environment of the space, and predicts the environment of the space after control based on the acquired information. [5] The environmental control system according to any one of [1] to [4], wherein the determining means determines whether or not to perform control with the determined control content in accordance with the comparison. [6] An environmental control method that is a method for operating an environmental control system that controls the environment of a space, comprising: a distribution acquisition step of acquiring control distribution information indicating a probability distribution of the content of control and target distribution information indicating a probability distribution of the environment of a space that is a target of control; a prediction determination step of determining a plurality of control contents to be used for predicting the environment of the space according to the control distribution information acquired in the distribution acquisition step; a prediction step of predicting a post-control environment of a space when control is performed using each of the plurality of control contents determined in the prediction determination step; a determination step of comparing the environment of the space after control for each of the plurality of control contents predicted in the prediction step with a probability distribution of the environment indicated by the target distribution information acquired in the distribution acquisition step, and determining the control contents for the space in accordance with the comparison; An environmental control method comprising: [Explanation of symbols]

[0084] 10...environmental control system, 20...air conditioner, 30...sensor, 11...distribution acquisition unit, 12...prediction determination unit, 13...prediction unit, 14...determination unit.

Claims

1. An environmental control system for controlling the environment of a space, a distribution acquisition means for acquiring control distribution information indicating a probability distribution of the content of control and target distribution information indicating a probability distribution of the environment of a space that is a target of control; a prediction determining means for determining a plurality of control contents to be used for predicting the environment of the space in accordance with the control distribution information acquired by the distribution acquiring means; a prediction means for predicting a post-control environment of a space when control is performed using each of the plurality of control contents determined by the prediction determination means; a determination means for comparing the environment of the space after control for each of the plurality of control contents predicted by the prediction means with a probability distribution of the environment indicated by the target distribution information acquired by the distribution acquisition means, and determining the control contents for the space in accordance with the comparison; An environmental control system comprising:

2. 2. The environmental control system according to claim 1, wherein the content of the control includes a plurality of items.

3. 3. The environmental control system according to claim 1, wherein the environment of the space to be controlled is at least one of temperature, humidity, wind speed, and particle concentration.

4. 3. The environmental control system according to claim 1, wherein the prediction means acquires information indicating the measurement results of the environment of the space, and predicts the environment of the space after control based on the acquired information.

5. 3. The environmental control system according to claim 1, wherein the determining means determines whether or not to perform control according to the determined control content in accordance with the comparison.

6. An environmental control method that is an operating method of an environmental control system that controls the environment of a space, comprising: a distribution acquisition step of acquiring control distribution information indicating a probability distribution of the content of control and target distribution information indicating a probability distribution of the environment of a space that is a target of control; a prediction determination step of determining a plurality of control contents to be used for predicting the environment of the space according to the control distribution information acquired in the distribution acquisition step; a prediction step of predicting a post-control environment of a space when control is performed using each of the plurality of control contents determined in the prediction determination step; a determination step of comparing the environment of the space after control for each of the plurality of control contents predicted in the prediction step with a probability distribution of the environment indicated by the target distribution information acquired in the distribution acquisition step, and determining the control contents for the space in accordance with the comparison; An environmental control method comprising:

Images