Prediction device and prediction method

Abstract

One embodiment of the present invention is a prediction device comprising a control unit that executes a room temperature prediction model, which is a mathematical model for predicting the temperature of a target space to be cooled or heated by an air conditioning device, on the basis of thermal environment information, target flow-of-people data, clothing information, and clothing statistical information, wherein: the thermal environment information indicates information pertaining to the thermal environment of the target space; the target flow-of-people data indicates the flow of people in the target space; the clothing information indicates the clothing of the people who are in the target space; and the clothing statistical information indicates statistical data that is obtained from a set of some or all pieces of past clothing information and that pertains to the clothing of people who were in the target space.

Description

Prediction Apparatus and Prediction Method

[0001] The present invention relates to a prediction apparatus and a prediction method.

[0002] There is a technique of predicting room temperature and controlling the temperature of a space to be controlled based on the prediction result (see Patent Document 1).

[0003] Japanese Unexamined Patent Application Publication No. 2019-60514

[0004] However, the control result may not always be appropriate. Specifically, as a result of control, there have been cases where the temperature becomes too hot or too cold compared to the expected temperature. This problem arises due to the low accuracy of room temperature prediction.

[0005] In view of the above circumstances, an object of the present invention is to provide a technique for predicting room temperature with higher accuracy.

[0006] One aspect of the present invention is a prediction apparatus comprising a control unit that executes a room temperature prediction model, the room temperature prediction model being a mathematical model that predicts the temperature of a target space, which is a space cooled or heated by an air conditioning apparatus, based on thermal environment information, target pedestrian flow data, clothing information, and clothing statistical information, wherein the thermal environment information is information indicating information on the thermal environment of the target space, the target pedestrian flow data is data indicating the flow of people in the target space, the clothing information is information indicating the clothing of people in the target space, and the clothing statistical information is statistical data obtained from a set of part or all of past clothing information and is information indicating statistical data of clothing of people who have been in the target space.

[0007] One aspect of the present invention is a prediction method comprising a control step in which a computer executes a room temperature prediction model, which is a mathematical model that predicts the temperature of a target space, which is a space to be cooled or heated by an air conditioning device, based on thermal environment information, target pedestrian flow data, clothing information, and clothing statistical information, wherein the thermal environment information is information indicating information about the thermal environment of the target space, the target pedestrian flow data is data indicating the flow of people in the target space, the clothing information is information indicating the clothing of people in the target space, and the clothing statistical information is statistical data obtained from a set of some or all of the past clothing information, indicating statistical data on the clothing of people who were in the target space.

[0008] This invention makes it possible to predict room temperature with higher accuracy.

[0009] An explanatory diagram illustrating the control system of the embodiment. A diagram showing an example of the hardware configuration of the control device in the embodiment. A flowchart showing an example of the processing flow executed by the control system of the embodiment.

[0010] (Embodiment) Figure 1 is an explanatory diagram illustrating a control system 100 of an embodiment. The control system 100 comprises at least an air conditioning device 1 and a control device 2.

[0011] Air conditioning equipment 1 is, for example, an air conditioner, a device that cools or heats a target space (hereinafter referred to as "target space"). In most cases, the target space is the space in which air conditioning equipment 1 is installed. However, if air conditioning equipment 1 is a device that blows out cold or hot air and is connected to another space by a hose or pipe, the target space is that other space.

[0012] The control device 2 includes a control unit 21 which is equipped with a processor 91 such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), or NPU (Neural Network Processing Unit) connected by a bus, and a memory 92, and executes a program. The control device 2 functions as a device comprising the control unit 21, interface unit 22, and storage unit 23 by executing the program.

[0013] The control unit 21 performs room temperature prediction processing. The control unit 21 may further perform air conditioning control processing. Note that the control device 2 does not necessarily have to perform air conditioning control processing. Air conditioning control processing may be performed by a device other than the control device 2. In this case, the results of the room temperature prediction processing may be sent to the device performing the air conditioning control processing manually or by communication.

[0014] <Room Temperature Prediction Process> The room temperature prediction process is the process of running a room temperature prediction model. The room temperature prediction model predicts the temperature of the target space (i.e., room temperature) based on at least thermal environment information, target pedestrian flow data, clothing information, and clothing statistics information.

[0015] Thermal environment information refers to information about the thermal environment of the target space. Target pedestrian flow data refers to data showing the flow of people in the target space. Since target pedestrian flow data refers to data showing the flow of people in the target space, it can also be said to be pedestrian flow data for the target space.

[0016] Clothing information refers to information that indicates the clothing of people in the target space. Clothing statistics information is statistical data obtained from some or all of the past clothing information, and refers to statistical data on the clothing of people who were in the target space (hereinafter referred to as "clothing statistics data").

[0017] The term "past" refers to the time before the prediction start time, which is defined as the time when the execution of this room temperature prediction model begins. The temperature predicted by the room temperature prediction model is used to control the air conditioning equipment 1, and is used to operate the air conditioning equipment 1 so that the temperature of the target space reaches a predetermined temperature at a predetermined time (hereinafter referred to as the "control target time").

[0018] Therefore, the temperature predicted by the room temperature prediction model is, more specifically, the temperature of the target space at a predetermined time defined as the target time. The predetermined time defined as the target time may be, for example, a predetermined time after the prediction by the room temperature prediction model has finished. This predetermined time may be specified in minutes, such as one minute later, or in hours, such as two hours later. Alternatively, the predetermined time defined as the target time may be a time that satisfies the condition of a predetermined timing within the period until the room temperature prediction model is run again.

[0019] The predetermined timing may be, for example, the middle of the period between the end of one room temperature prediction model's execution and the start of the next room temperature prediction model's execution. The predetermined timing may be, for example, a timing within the period between the end of one room temperature prediction model's execution and the start of the next room temperature prediction model's execution, and may also be a predetermined time after the end of one room temperature prediction model's execution.

[0020] The control target time may be entered by the user each time the room temperature prediction process is performed, or it may be pre-recorded in the storage unit 23, etc., without the user needing to enter it each time the room temperature prediction process is executed. In other words, the room temperature prediction model has a predetermined definition of what time of day the room temperature will be predicted. Of course, the control target time recorded in this storage unit 23 may be overwritten by the user.

[0021] The room temperature prediction model can be any mathematical model that predicts the room temperature of a target space based on at least thermal environment information, target pedestrian flow data, clothing information, and clothing statistics.

[0022] The room temperature prediction model can be obtained in any way, for example, by machine learning. This machine learning may be supervised learning that uses room temperature as the ground truth data (or labels). Such supervised learning may be supervised learning using LSTM (Long Short Term Memory), supervised learning using ARIMA (AutoRegressive Integrated Moving Average), or supervised learning using linear regression.

[0023] Thermal environment information may, for example, indicate the temperature measured by a thermometer installed in the target space. Thermal environment information may also indicate the humidity measured by a hygrometer installed in the target space. Thermal environment information may also indicate the setting value of air conditioning equipment 1. Thermal environment information may also indicate information about heat source devices such as lighting equipment and electrical equipment present in the target space. Thermal environment information may also indicate layout data of the facility including the target space. Thermal environment information may also indicate the physical characteristics of the target space. Physical characteristics may include, for example, the thermal insulation of the wall separating the target space from other spaces. Physical characteristics may also include, for example, solar radiation. Thermal environment information may also indicate the temperature of the space outside the target space (i.e., the outside temperature).

[0024] The target pedestrian flow data (pedestrian flow data) may be information that shows pedestrian flow predicted using a mathematical model that predicts pedestrian flow based on the location of people (hereinafter referred to as the "pedestrian flow prediction model"). More specifically, the target pedestrian flow data may show the results of predicting pedestrian flow in the target space based on the location of each person present in the target space using the pedestrian flow prediction model.

[0025] The pedestrian flow prediction model may utilize well-known techniques for predicting pedestrian flow based on people's location information. The pedestrian flow prediction model may, for example, be obtained through machine learning.

[0026] Clothing information indicates, for example, whether each person in the target space is lightly dressed, heavily dressed, or neither lightly nor heavily dressed. Clothing information may also indicate, for example, whether they are wearing long sleeves or not. Clothing information may also indicate, for example, whether they are wearing gloves or not. Clothing information may also indicate, for example, whether they are wearing a scarf or not. Clothing information may also indicate, for example, the clothing index (Clo value).

[0027] Clothing information can be obtained, for example, by image analysis of an image obtained by photographing the target space. Therefore, clothing information may be obtained by a device that obtains clothing information by performing image analysis on an image obtained by photographing the target space. Alternatively, clothing information may be obtained by a device that photographs the target space and performs image analysis on the image obtained from that photograph. Any well-known technique may be used for obtaining clothing information by image analysis of an image obtained by photographing the target space.

[0028] Furthermore, clothing information may be the result of people in the target space self-reporting their clothing. For example, if a person in the target space reports their clothing to a designated server using a device such as a smartphone, the clothing information may represent that reported content. Alternatively, clothing information may be the result of the administrator of the target space converting information about the clothing reported by a person in the target space on paper into electronic data.

[0029] Clothing statistics may include statistical data corresponding to the content of the clothing information, for example, the percentage of people wearing light clothing or the percentage of people wearing heavy clothing. Clothing statistics may also include, for example, the percentage of people wearing long sleeves, the percentage of people wearing gloves, or the percentage of people wearing scarves. Clothing statistics may also include, for example, the average clothing level as statistical data related to the clothing index. The average clothing level may be shown in three stages, for example, light clothing, normal clothing, and heavy clothing, or it may be shown as a numerical index.

[0030] <Air Conditioning Control Processing> Air conditioning control processing is the process of controlling the operation of air conditioning equipment 1. More specifically, air conditioning control processing is the process of controlling the operation of air conditioning equipment 1 using the temperature predicted by the room temperature prediction processing so that the temperature of the target space becomes a predetermined temperature. Even more specifically, air conditioning control processing is the process of controlling the operation of air conditioning equipment 1 using the temperature predicted by the room temperature prediction processing so that the temperature of the target space at the control target time becomes a predetermined temperature. Here, an example of processing that uses the temperature predicted by the room temperature prediction processing so that the temperature of the target space becomes a predetermined temperature will be explained.

[0031] Let's assume that the room temperature prediction process predicts that "the room temperature will rise by K°C in the next S hours." S and K are both positive real numbers. In this case, the air conditioning control process instructs the air conditioning unit 1 to "reduce the airflow of the cooling system." The prediction uses data from a room temperature sensor, an outside temperature sensor, clothing statistics, and user count information, and the temperature change in the target space is modeled based on the obtained data. The modeling process is performed by the control unit 21 of the control device 2.

[0032] Next, the control unit 21 determines the values ​​of the operating parameters of the air conditioning unit 1 (for example, airflow, airflow direction, temperature setting, operating mode, etc.). In determining the parameter values, the control unit 21 first recalculates the target temperature based on the room temperature prediction result and the user's clothing statistics. If the target temperature is higher than the current room temperature, the control unit 21 determines the values ​​of the operating parameters so that the air conditioning unit 1 operates in heating mode. If the target temperature is lower than the current room temperature, the control unit 21 determines the values ​​of the operating parameters so that the air conditioning unit 1 operates in cooling mode. If the difference from the target temperature is smaller than a predetermined difference, the control unit 21 determines the values ​​of the operating parameters so that the air conditioning unit 1 makes predetermined fine adjustments based on the above difference, such as reducing the airflow.

[0033] After the adjustment, the control unit 21 transmits the determined operating parameter values ​​to the air conditioning unit 1. Upon receiving the operating parameter values, the air conditioning unit 1 performs operations according to those values. In this way, the control unit 21 controls the operation of the air conditioning unit 1. In this manner, the air conditioning control process achieves both user comfort and energy-saving performance by dynamically adjusting the operating parameters of the air conditioning unit 1 based on the predicted temperature.

[0034] <Effects> The room temperature prediction model predicts the temperature of a target space based on at least thermal environment information, target pedestrian flow data, clothing information, and clothing statistics. Therefore, it can predict the temperature of a target space with higher accuracy than mathematical models that predict the temperature of a target space without using thermal environment information, mathematical models that predict the temperature of a target space without using target pedestrian flow data, mathematical models that predict the temperature of a target space without using clothing information, or mathematical models that predict the temperature of a target space without using clothing statistics. Consequently, the control device 2 that executes the room temperature prediction model to predict the temperature of a target space can predict the temperature of the target space with higher accuracy.

[0035] Furthermore, since the temperature of the target space can be predicted with high accuracy by running the room temperature prediction model, excessive air conditioning operation is suppressed, and energy consumption is reduced.

[0036] <Example of Hardware Configuration of Control Device 2> Figure 2 shows an example of the hardware configuration of the control device 2 in the embodiment. The control device 2 includes a control unit 21 and executes a program. By executing the program, the control device 2 functions as a device comprising the control unit 21, an interface unit 22 equipped with a communication interface 121, and a storage unit 23.

[0037] More specifically, the processor 91 reads the program stored in the storage unit 23 and stores the read program in the memory 92. By executing the program stored in the memory 92, the processor 91 functions as a device comprising a control unit 21, an interface unit 22, and a storage unit 23.

[0038] The control unit 21 controls the operation of each functional unit of the control device 2. The control unit 21 performs, for example, room temperature prediction processing as described above. The control unit 21 may also perform, for example, air conditioning control processing as described above. The control unit 21 controls, for example, the operation of the interface unit 22. The control unit 21 controls the operation of the interface unit 22 to obtain, for example, information acquired by the interface unit 22. The control unit 21 obtains, for example, information stored in the storage unit 23. Specifically, the process of obtaining information stored in the storage unit 23 is a read operation.

[0039] The interface unit 22 is configured to include a communication interface for connecting the control device 2 to an external device. The interface unit 22 communicates with the external device via wired or wireless means.

[0040] The external device may be, for example, a device that transmits information used in predictions by a room temperature prediction model.

[0041] Therefore, the external device may be, for example, a device that transmits thermal environment information. The interface unit 22 acquires thermal environment information by communicating with the device that transmits thermal environment information. The device that transmits thermal environment information may be, for example, a thermometer. In this case, the interface unit 22 acquires information indicating the temperature measured by the thermometer as thermal environment information. The device that transmits thermal environment information may be, for example, a hygrometer. In this case, the interface unit 22 acquires information indicating the humidity measured by the hygrometer as thermal environment information. The source of the thermal environment information may be a server that stores thermal environment information. In this case, the interface unit 22 acquires the thermal environment information recorded on the server.

[0042] The external device may be, for example, a device that is the source of information used for prediction by a room temperature prediction model, and thus the external device may be, for example, a device that is the source of target pedestrian flow data. The interface unit 22 acquires the target pedestrian flow data through communication with the device that is the source of the target pedestrian flow data. The device that is the source of the target pedestrian flow data may be, for example, a device that executes a pedestrian flow prediction model. In such a case, the interface unit 22 acquires information indicating the pedestrian flow predicted by the device that executes this pedestrian flow prediction model. The device that is the source of the target pedestrian flow data does not necessarily need to execute the pedestrian flow prediction model by itself, and may be a device that stores pedestrian flow information predicted by another device using the pedestrian flow prediction model. Furthermore, the device that is the source of the target pedestrian flow data may be a device that stores manually predicted pedestrian flow information.

[0043] The external device may be, for example, a device that is the source of information used for prediction by a room temperature prediction model, and thus the external device may be, for example, a device that is the source of clothing information. The interface unit 22 acquires clothing information through communication with the device that is the source of the clothing information. The device that is the source of the clothing information may be, for example, a device that obtains clothing information by performing image analysis on an image obtained by capturing an image of the target space.

[0044] The device that is the source of the clothing information may be, for example, a device that captures an image of the target space and performs image analysis on the image obtained by the capturing to obtain clothing information. The interface unit 22 acquires the clothing information acquired by the device that is the source of the clothing information. The device that is the source of the clothing information does not necessarily need to perform image analysis by itself, and may be a device that stores clothing information obtained by another device through image analysis. Furthermore, the device that is the source of the clothing information may be a device that stores manually acquired clothing information. Acquiring clothing information manually means visually observing the target space and creating clothing information manually.

[0045] The external device may be, for example, a device that is a transmission source of information used for prediction by a room temperature prediction model, and thus may be, for example, a device that is a transmission source of clothing statistical information. The interface unit 22 acquires the clothing statistical information through communication with the device that is the transmission source of the clothing statistical information. The device that is the transmission source of the clothing statistical information may be, for example, a device that generates the clothing statistical information by performing statistical analysis based on a set of clothing information.

[0046] In such a case, the interface unit 22 acquires the clothing statistical information obtained by the device that is the transmission source of the clothing information. The device that is the transmission source of the clothing statistical information does not necessarily need to perform statistical analysis by itself, and may be a device that stores clothing statistical information obtained through statistical analysis by another device. Further, the device that is the transmission source of the clothing statistical information may be a device that stores manually acquired clothing statistical information. Acquiring clothing statistical information manually means performing statistical analysis manually based on a set of clothing information.

[0047] Note that the clothing statistical information does not necessarily need to be acquired from an external device, and may be stored in the storage unit 23 in advance.

[0048] The interface unit 22 may be configured to include input devices such as a mouse, a keyboard, and a touch panel. The interface unit 22 may be configured as an interface that connects these input devices to the control device 2. As described above, the input devices of the interface unit 22 accept input of various types of information to the control device 2 via wired or wireless connection. Note that information does not necessarily need to be input to the communication interface of the interface unit 22, and may be input to an input device of the interface unit 22. That is, the information may be input manually.

[0049] The interface unit 22 outputs various types of information, for example. The interface unit 22 is comprised of a display device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, or an organic EL (Electro-Luminescence) display, as well as a speaker. The interface unit 22 may be configured as an interface for connecting these display devices or speakers to the control device 2. Therefore, the interface unit 22 may output information input to its input device as an image or sound, for example.

[0050] The storage unit 23 is configured using a computer-readable storage medium (non-transitory computer-readable recording medium), such as a magnetic hard disk drive or a semiconductor storage device. The storage unit 23 may reside, for example, on the cloud.

[0051] The memory unit 23 stores various information related to the control device 2. The memory unit 23 stores various information generated by the operation of the control unit 21, for example. Therefore, the memory unit 23 may store clothing information obtained from an external device via the interface unit 22, for example.

[0052] Therefore, the control unit 21 may, for example, read a set of clothing information from the storage unit 23 and perform statistical analysis to obtain clothing statistics. In other words, the control unit 21 may perform clothing statistical analysis on the set of clothing information stored in the storage unit 23. Clothing statistical analysis is a process that performs statistical analysis on a set of information to be processed to obtain clothing statistics.

[0053] Figure 3 is a flowchart showing an example of the processing flow executed by the control system 100 of this embodiment.

[0054] The control unit 21 acquires information used for prediction by the room temperature prediction model (step S101). The information used for prediction by the room temperature prediction model includes at least thermal environment information, target pedestrian flow data, clothing information, and clothing statistics information.

[0055] In step S101, the acquisition of clothing statistics information is a process in which the control unit 21 reads the clothing statistics information from the storage unit 23 if the clothing statistics information is already stored in the storage unit 23. Also, if the control unit 21 is executing a clothing statistics analysis process, the acquisition of clothing statistics information in step S101 is a process in which the clothing statistics information is obtained by executing the clothing statistics analysis process.

[0056] The control unit 21 does not necessarily need to obtain target pedestrian flow data via the interface unit 22. The control device 2 may receive information used for prediction by the pedestrian flow prediction model from an external device or manually via the interface unit 22. The information used for prediction by the pedestrian flow prediction model is, for example, the location information of people in the target space. In this case, the control unit 21 itself may execute the pedestrian flow prediction model to obtain the target pedestrian flow data. In this case, the acquisition of target pedestrian flow data in step S101 means the process of obtaining the target pedestrian flow data by the control unit 21 itself executing the pedestrian flow prediction model.

[0057] The control unit 21 does not necessarily need to obtain clothing information via the interface unit 22. The control device 2 may receive information used for image analysis to obtain clothing information from an external device or manually via the interface unit 22. The information used for image analysis to obtain clothing information is, for example, image data obtained by photographing the target space. In this case, the control unit 21 itself may perform the image analysis to obtain clothing information. In this case, the acquisition of clothing information in step S101 means the process of obtaining clothing information by the control unit 21 itself performing the image analysis.

[0058] Following step S101, the control unit 21 performs room temperature prediction processing (step S102). Next, the control unit 21 controls the operation of the air conditioning equipment 1 using the temperature predicted in step S102 so that the temperature of the target space at the target control time becomes a predetermined temperature (step S103). In other words, the control unit 21 performs air conditioning control processing.

[0059] The control device 2 configured in this way includes a control unit 21 that performs room temperature prediction processing. Therefore, as described in <Effects Produced>, the control device 2 can predict the temperature of the target space with higher accuracy.

[0060] (Modification) The thermal environment information used in the room temperature prediction model may be, for example, information on the thermal environment of the target space at a time prior to the prediction start time, where the difference from the prediction start time is within a first predetermined difference.

[0061] The target pedestrian flow data used in the room temperature prediction model may, for example, be pedestrian flow information obtained based on the location information of people in the target space at a time prior to the prediction start time, where the difference from the prediction start time is within the second predetermined difference. The first predetermined difference and the second predetermined difference may be the same or different.

[0062] The clothing information used in the room temperature prediction model may, for example, be information about the clothing of a person in the target space at a time before the prediction start time, where the difference from the prediction start time is within a third predetermined difference. The third predetermined difference may be the same as or different from the first predetermined difference. Also, the third predetermined difference may be the same as or different from the second predetermined difference.

[0063] Furthermore, the statistical data shown in the clothing statistics information (i.e., clothing statistics data) may be shown for predetermined periods. The predetermined period may be, for example, every hour, every morning and every afternoon, every day of the week, every month, every season, every date, or every combination of some or all of these. When the clothing statistics information shows clothing statistics data for predetermined periods, the room temperature prediction model makes predictions using the clothing statistics data for the period including the control target time.

[0064] <Body Temperature Information> The room temperature prediction model may also make predictions based on body temperature information. That is, body temperature information may be included in the information used for prediction by the room temperature prediction model. Body temperature information is information that indicates the body temperature of a person in the target space. Body temperature information may be obtained, for example, by a thermographic camera. Body temperature information may also be obtained, for example, by having a person in the target space measure their body temperature using a body temperature sensing sensor such as a thermometer.

[0065] When training a room temperature prediction model that also makes predictions based on temperature information, the temperature information is used as input data for the room temperature prediction model.

[0066] The temperature information used in the room temperature prediction model may, for example, be the body temperature of a person in the target space at a time prior to the prediction start time, where the difference from the prediction start time is within the fourth predetermined difference. The fourth predetermined difference may be the same as or different from the first predetermined difference. Furthermore, the fourth predetermined difference may be the same as or different from the second predetermined difference. Furthermore, the fourth predetermined difference may be the same as or different from the third predetermined difference.

[0067] <Perceived Temperature Declaration Information> The room temperature prediction model may also make predictions based on perceived temperature declaration information. In other words, the information used in the prediction by the room temperature prediction model may include perceived temperature declaration information. Perceived temperature declaration information is information indicating the perceived temperature of a person who is in the target space, as declared by that person. For example, if a person in the target space declares their perceived temperature to a designated server using a terminal such as a smartphone, the perceived temperature declaration information may indicate the content of that declaration. In addition, the perceived temperature declaration information may be the result of the administrator of the target space converting information on the perceived temperature of a person in the target space who declared it on paper into electronic data.

[0068] Self-reported perceived temperature includes information such as "hot," "cold," or "just right."

[0069] When training a room temperature prediction model that also makes predictions based on perceived temperature reports, the perceived temperature reports are used as input data for the room temperature prediction model.

[0070] Furthermore, the perceived temperature report information used in the room temperature prediction model may be, for example, information on the perceived temperature at a time reported by a person in the target space at a time before the prediction start time, where the difference from the prediction start time is within the fifth predetermined difference. The fifth predetermined difference may be the same as or different from the first predetermined difference. Also, the fifth predetermined difference may be the same as or different from the second predetermined difference. Also, the fifth predetermined difference may be the same as or different from the third predetermined difference. Also, the fifth predetermined difference may be the same as or different from the fourth predetermined difference.

[0071] <Perceived Temperature Statistics> Room temperature prediction models may also make predictions based on perceived temperature statistics. In other words, the information used in predictions by room temperature prediction models may include perceived temperature statistics.

[0072] The perceived temperature statistics are statistical data obtained from a collection of some or all past perceived temperature reports, and represent statistical data of perceived temperature reported by people who were in the target space (hereinafter referred to as "perceived temperature statistics").

[0073] Furthermore, the statistical data shown by the perceived temperature statistics information (i.e., perceived temperature statistics data) may be shown at predetermined intervals. The predetermined interval may be, for example, every hour, every morning and every afternoon, every day of the week, every month, every season, every date, or every combination of some or all of these. When the perceived temperature statistics information shows perceived temperature statistics data at predetermined intervals, the room temperature prediction model makes predictions using the perceived temperature statistics data for the period including the control target time.

[0074] <Congestion Level> The room temperature prediction model may infer the congestion level in the target space based on the target pedestrian flow data, and predict the temperature of the target space based on the inferred congestion level. For example, if the target pedestrian flow data indicates that there are 200 people in the target space at 10:00 AM and is predicted to be 300 people at 2:00 PM, the model may infer temperature information corresponding to that congestion level and predict the temperature of the target space based on the inferred temperature information.

[0075] <Clothing Advice Generation Process> The control unit 21 may also perform clothing advice information generation processing. Clothing advice information generation processing is the process of executing the clothing advice information generation model. The clothing advice information generation model is a mathematical model that generates information (hereinafter referred to as "clothing advice information") that provides advice on clothing, such as "dress warmly" or "dress lightly," based on the temperature predicted by the room temperature prediction model. In other words, the clothing advice information generation model is a mathematical model that generates clothing advice information based on temperature.

[0076] The data format of the generated clothing advice information may be a predetermined format, for example, text data, image data, or audio data. The predetermined format may be a format determined according to the device that outputs the advice indicated by the clothing advice information. For example, if the advice is displayed on a display device, the predetermined format may be text data or image data, and if the advice is output through a speaker, the predetermined format may be audio data.

[0077] The control unit 21 may control the operation of a predetermined output device to output the clothing advice information generated in the clothing advice information generation process. The predetermined output device may be, for example, a display device provided by the interface unit 22, or a speaker.

[0078] If the designated output device is a display device, the display device, under the control of the control unit 21, displays the advice indicated by the clothing advice information in text or images. If the designated output device is a speaker, the speaker, under the control of the control unit 21, outputs the advice indicated by the clothing advice information in voice.

[0079] The clothing advice information generation model can be obtained in any way, for example, by machine learning. This machine learning may be supervised learning that uses clothing advice information as ground truth data (or labels). Such supervised learning may be supervised learning using LSTM (Long Short Term Memory), supervised learning using ARIMA (AutoRegressive Integrated Moving Average), or supervised learning using linear regression.

[0080] The control device 2 may be implemented using multiple information processing devices connected to each other via a network. In this case, each process executed by the control unit 21 may be distributed among multiple information processing devices. In this case, for example, the room temperature prediction process and the air conditioning control process may be executed by different information processing devices.

[0081] Furthermore, all or part of the functions of the control system 100 may be implemented using hardware such as ASICs (Application Specific Integrated Circuits), PLDs (Programmable Logic Devices), or FPGAs (Field Programmable Gate Arrays). The program may be recorded on a computer-readable recording medium. Computer-readable recording media include, for example, portable media such as flexible disks, magneto-optical disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into computer systems. The program may also be transmitted via a telecommunications line.

[0082] Note that control device 2 is an example of a prediction device.

[0083] While embodiments of this invention have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments and includes designs and the like that do not depart from the spirit of this invention.

[0084] 100...Control system, 1...Air conditioning equipment, 2...Control device, 21...Control unit, 22...Interface unit, 23...Storage unit, 91...Processor, 92...Memory

Claims

1. A prediction device comprising: a control unit that executes a room temperature prediction model, which is a mathematical model that predicts the temperature of a target space, which is a space to be cooled or heated by an air conditioning device, based on thermal environment information, target pedestrian flow data, clothing information, and clothing statistical information, wherein the thermal environment information is information indicating information about the thermal environment of the target space, the target pedestrian flow data is data indicating the flow of people in the target space, the clothing information is information indicating the clothing of people in the target space, and the clothing statistical information is statistical data obtained from a set of some or all of the past clothing information, indicating statistical data on the clothing of people who were in the target space.

2. The prediction device according to claim 1, wherein the control unit further executes a clothing advice information generation model, and the clothing advice information generation model generates clothing advice information, which is information indicating advice on clothing, based on the temperature predicted by the room temperature prediction model.

3. The prediction device according to claim 1 or 2, wherein the room temperature prediction model further predicts the temperature of the target space based on subjective temperature declaration information, which is information indicating the subjective temperature reported by a person in the target space.

4. A prediction method comprising: a control step in which a computer executes a room temperature prediction model, which is a mathematical model that predicts the temperature of a target space, which is a space to be cooled or heated by an air conditioning device, based on thermal environment information, target pedestrian flow data, clothing information, and clothing statistical information, wherein the thermal environment information is information indicating information about the thermal environment of the target space, the target pedestrian flow data is data indicating the flow of people in the target space, the clothing information is information indicating the clothing of people in the target space, and the clothing statistical information is statistical data obtained from a set of some or all of the past clothing information, indicating statistical data on the clothing of people who were in the target space.

Images