air conditioning unit

Abstract

To provide air conditioning equipment that easily achieves both improved comfort and improved energy efficiency. [Solution] An air conditioning device 10 that adjusts a first temperature which is the temperature of a target space, is characterized by comprising: a prediction unit 103 that determines a predicted first temperature which is the first temperature after a predetermined time based on the first temperature, a second temperature which is the temperature outside the target space, the set temperature of the air conditioning device 10, and the number of people staying in the target space; an accuracy unit 105 that determines the prediction accuracy based on the predicted first temperature and the first temperature after a predetermined time; and a control unit 106 that controls the air conditioner so that the air conditioning index falls within a predetermined first range, wherein if the prediction accuracy exceeds a predetermined threshold, the control unit 106 controls the air conditioner so that the air conditioning index falls within the first range, and if the prediction accuracy does not exceed the threshold, the control unit 106 controls the air conditioning index to fall within a second range which is narrower than the first range, or controls without using the air conditioning index.

Description

Technical Field

[0001] The present invention relates to an air conditioner.

Background Art

[0002] In recent years, in order to improve the comfort during occupancy, a method of calculating a set temperature by a machine and controlling an air conditioner based on the calculated set temperature is known. As an example of the control of the air conditioner described above, for example, Patent Document 1 is disclosed.

[0003] In the control method of the air conditioner of Patent Document 1, at least an optimal air conditioning control method considering the housing heat retention capacity is calculated. By performing the calculated air conditioning control method, the comfort during occupancy is improved.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] When performing the control as described above, the air conditioner performs a certain control based on the calculated air conditioning control method. However, when the accuracy of the information considered in the calculated air conditioning control method is low, there is a problem that it is difficult to improve the comfort.

[0006] The present invention has been made to solve the above problems, and an object thereof is to provide an air conditioner that can easily achieve both improvement in comfort and improvement in energy saving effect.

Means for Solving the Problems

[0007] To achieve the above object, the present invention provides the following means. An air conditioning system according to one aspect of the present invention is an air conditioning system that adjusts a first temperature which is the temperature of a target space, and is characterized by comprising: a prediction unit that determines a predicted first temperature which is the first temperature after a predetermined time based on the first temperature, a second temperature which is the temperature outside the target space, the set temperature of the air conditioning system, and the number of people staying in the target space; an accuracy unit that determines the prediction accuracy based on the predicted first temperature and the first temperature after the predetermined time; and a control unit that controls the air conditioning system so that the air conditioning index falls within a predetermined first range, wherein if the prediction accuracy exceeds a predetermined threshold, the control unit controls the air conditioning index to fall within the first range, and if the prediction accuracy does not exceed the threshold, the control unit controls the air conditioning index to fall within a second range which is narrower than the first range, or controls the system without using the air conditioning index.

[0008] According to the first aspect of the present invention, different control can be performed depending on whether the calculated prediction accuracy exceeds a threshold (hereinafter also referred to as high accuracy) or whether the prediction accuracy does not exceed a threshold (hereinafter also referred to as low accuracy). By performing different control in high and low accuracy cases, the air conditioning system can easily improve comfort. Furthermore, it is easier to achieve both improved comfort and improved energy saving effects.

[0009] The first range refers to the range of air conditioning indicators that are most likely to improve comfort. Specifically, the first range is preferably one where the PMV value is within the range of -0.5 to +0.5. The second range is a narrower range of air conditioning indicators than the first range, and is preferably a range of air conditioning indicators that easily improves comfort. More specifically, the second range is preferably a range that easily achieves both improved comfort and improved energy saving effects. In the embodiments described below, the second range is preferably a range narrower than the range of PMV values ​​from -0.5 to +0.5. Note that the second range may be a range other than those described above.

[0010] The air conditioning index refers to the PMV value. Details about the air conditioning index and PMV value will be explained later. In the first embodiment of the above invention, it is preferable that the prediction unit estimates the predicted first temperature by inputting the first temperature, the second temperature, the set temperature of the air conditioning unit, and the number of people staying in the target space into a trained learning model that has been trained by machine learning to estimate the predicted first temperature after a predetermined time in the target space.

[0011] Thus, by using machine learning to estimate the first predicted temperature, the prediction unit can estimate the first predicted temperature with higher accuracy compared to estimating the first predicted temperature without machine learning. By estimating the first predicted temperature with high accuracy, it is easier to achieve both improved comfort and improved energy saving effects.

[0012] In the first embodiment of the above invention, it is preferable that the control unit controls the air conditioning indicator to fall within the first range or the second range by changing at least one of the start time of the air conditioning unit, the stop time of the air conditioning unit, and the set temperature of the air conditioning unit.

[0013] In this way, the control unit can change the start time of the air conditioning unit, the stop time of the air conditioning unit, and the set temperature of the air conditioning unit (hereinafter also referred to as variables, etc.). By changing the variables, etc., it is easy to diversify the control content and to achieve both improved comfort and improved energy saving effects.

[0014] An air conditioning system according to one aspect of the present invention is an air conditioning system that adjusts a first temperature which is the temperature of a target space, and is characterized by comprising: a prediction unit that determines a predicted first temperature which is the first temperature after a predetermined time based on the first temperature, a second temperature which is the temperature outside the target space, the set temperature of the air conditioning system, and the number of people staying in the target space; an accuracy unit that determines the prediction accuracy based on the predicted first temperature and the first temperature after a predetermined time; and a control unit that controls the air conditioning system so that the air conditioning index falls within a predetermined first range, wherein if the prediction accuracy exceeds a predetermined threshold, the control unit controls the air conditioning system so that the air conditioning index falls within the first range, and if the prediction accuracy does not exceed the threshold, the control unit controls the set temperature of the air conditioning system.

[0015] According to the air conditioning system of the second embodiment of the present invention, control is possible that easily achieves both improved comfort and improved energy saving effects. The air conditioning system of this embodiment easily achieves both improved comfort and improved energy saving effects, and compared to the first embodiment, it is easier to control the system to prioritize comfort. [Effects of the Invention]

[0016] According to the air conditioning system of the present invention, it is possible to achieve control that easily balances improved comfort with improved energy saving effects. [Brief explanation of the drawing]

[0017] [Figure 1] This is a block diagram illustrating the configuration of an air conditioning system according to the first embodiment of the present invention. [Figure 2] This graph illustrates prediction accuracy. [Figure 3] This is a flowchart explaining the control process of an air conditioning system. [Modes for carrying out the invention]

[0018] [First Embodiment] Description of the structure The air conditioner 10 according to the first embodiment of the present invention will be described with reference to FIGS. 1 and 2. The air conditioner 10 of the present embodiment is composed of at least one or more cameras, measuring instruments, not shown, an air conditioner 50, and a control device 100, and is a device for achieving both improvement in the comfort of the target space and improvement in energy saving effect.

[0019] The air conditioner 50 has a configuration in which it sucks in the air in the target space, cools it, and supplies the cooled air to the target space. Specifically, it has a configuration in which heat exchange is performed between the air and the refrigerant by circulating the refrigerant between an outdoor unit, not shown, and cools the air. The outdoor unit has a configuration in which the refrigerant that has taken heat from the air releases the heat to the outside air.

[0020] The air conditioner 50 is provided with a heat exchange unit (not shown) that cools the air sucked in by heat exchange, and an air conditioner fan (not shown) that sends out the heat-exchanged air. A heat exchange unit and an air conditioner fan having a known configuration are used.

[0021] The air conditioner 50 preferably conditions the air so that the target space reaches the set temperature. In the present embodiment, the case where the air conditioner 50 is installed in a corporate office will be described. Note that the air conditioner 50 may be installed in a building other than an office. For example, it may be installed in a house or a classroom in a school. Also, the air conditioner 50 may be installed in a means of transportation other than a building.

[0022] The air conditioner 50 has a configuration that can be communicably connected to cameras, measuring instruments, not shown, and the control device 100. Specifically, it is preferable that the air conditioner 50 is controlled based on a signal transmitted from the control device 100.

[0023] The air conditioner 50 discharges conditioned air so that the temperature of the target space reaches the set temperature. In this embodiment, it is preferable that the temperature of the target space reaches the set temperature by the start of business hours for the company. In other words, it is preferable that the air conditioner 50 is operating before the start of business hours for the company. Furthermore, it is preferable that the conditioned air discharged before the target space reaches the set temperature is at a temperature lower than the set temperature.

[0024] The target space (hereinafter also referred to as the interior) is preferably the interior of a building that is subject to air conditioning. The control device 100 is an information processing device such as a server or computer that has a CPU (Central Processing Unit), ROM, RAM, input / output interfaces, etc. The control device 100 is connected to multiple air conditioners 50 and cameras in a manner that enables information communication. The number of air conditioners 50 and cameras connected to the control device 100 may be one or multiple.

[0025] As shown in Figure 1, the control device 100 stores a program in its ROM, etc., that causes the CPU, ROM, RAM, and input / output interface to work together to function as at least an acquisition unit 101, a storage unit 102, a prediction unit 103, a measurement unit 104, an accuracy unit 105, a control unit 106, and a calculation unit 107.

[0026] The acquisition unit 101 is connected to a camera (not shown) to enable communication and acquisition of captured image information and measurement information measured by measuring instruments (hereinafter also referred to as image information, etc.). The acquisition unit 101 has a configuration to acquire image information including the activity level, clothing amount, and number of occupants captured by the camera, as well as the room temperature and other information measured by measuring instruments.

[0027] The image information preferably includes information for estimating the activity level of occupants, the amount of clothing they are wearing, and the number of occupants in the room. The amount of clothing refers to parameters indicating the type and number of garments worn by the occupants. For example, it may include the sleeve length and the number of outer garments worn by the occupants. Note that the amount of clothing may also include information other than that mentioned above.

[0028] Activity level refers to information indicating the state of the occupants. Specifically, it is preferable that this includes information about the occupants' movement, such as whether they are walking or sitting. The measurement information should preferably include information for estimating the air conditioning indicators for occupants, as described later. For example, it is preferable to include indoor temperature, humidity, radiant temperature, and wind speed.

[0029] The number of occupants refers to the number of people staying in the target space. The number of occupants may be calculated from image information, or it may be calculated using a scheduling system. A scheduling system is a system that calculates the location and duration of stays of occupants by recording the room and time they will be staying in in advance using information and communication equipment.

[0030] The storage unit 102 is an information storage medium that has the function of storing information for estimating the room temperature. Preferably, the stored information includes image information acquired by the acquisition unit 101. The storage unit 102 may be a flash memory such as an SD memory card, or it may be another type of recording medium.

[0031] The prediction unit 103 has a configuration for estimating a first predicted temperature. In this embodiment, the prediction unit 103 has a configuration for estimating a first predicted temperature by inputting the information described below into the learning model.

[0032] The information input to the first temperature prediction learning model includes the first temperature, the second temperature, the set temperature of the air conditioner 50, and the number of people in the room. Note that information other than the above may also be input to the learning model. Furthermore, the learning model is a model that has been trained using machine learning. For machine learning, known learning methods can be used.

[0033] The first temperature refers to the indoor temperature of the target space. The second temperature is the temperature outside the target space. In this embodiment, this is the ambient temperature. It is preferable to use the ambient temperature measured by a measuring instrument.

[0034] The first predicted temperature is the room temperature estimated by the prediction unit 103 after a predetermined time has elapsed. In this embodiment, the first predicted temperature is described as the room temperature after one hour. Note that the predetermined time for the first predicted temperature may be other than one hour.

[0035] The calculation unit 107 has a configuration that calculates an air conditioning index (hereinafter also referred to as PMV value) based on image information, etc. The calculated PMV value is used in the control method of the control unit 106, which will be described later.

[0036] The PMV value is a parameter calculated using a comfort equation with six variables: indoor temperature, humidity, radiant temperature, wind speed, amount of clothing worn, and activity level. In this embodiment, the PMV value is a parameter calculated using a comfort equation with the following variables: indoor temperature, humidity, and radiant temperature included in the information on the indoor environment; wind speed included in the information on the outdoor environment; and amount of clothing worn and activity level included in the information on occupants. In addition to the above-mentioned elements, the number of occupants may also be used as an element in the calculation of the PMV value.

[0037] The PMV value ranges from -3 (cold) to +3 (hot). Statistically, approximately 95% of people feel comfortable when the PMV value is 0, and approximately 90% of people feel comfortable when the PMV value is between -0.5 and +0.5. It is preferable to control the air conditioning so that the PMV value falls within this predetermined range.

[0038] The accuracy unit 105 has a configuration for calculating the prediction accuracy of the predicted first temperature. In this embodiment, the prediction accuracy is calculated based on the predicted first temperature and the first temperature after a predetermined time has elapsed corresponding to the predicted first temperature. Specifically, it is preferable that the accuracy unit 105 calculates the prediction accuracy based on the estimated predicted first temperature after one hour and the first temperature obtained after one hour. Furthermore, the accuracy unit 105 has a configuration for determining whether the calculated prediction accuracy exceeds a predetermined threshold.

[0039] In this embodiment, the prediction accuracy is preferably expressed as the absolute value of the temperature difference between the estimated predicted first temperature and the first temperature measured at the time corresponding to the predicted first temperature. However, the prediction accuracy may be a value other than the one described above.

[0040] For example, in addition to the above, the prediction accuracy may also be a value that indicates the difference between the air-conditioned temperature discharged from the air conditioner 50 and a predetermined air-conditioned temperature. Specifically, it is preferable that it be a value that indicates the difference between the slope of line D and the slope of line E in Figure 2. Note that the prediction accuracy may also be a value other than the two examples above.

[0041] Figure 2 shows the energy consumption of the air conditioner 50. The vertical axis represents the air-conditioned temperature discharged from the air conditioner 50. The horizontal axis represents time. Vertical line A represents the start time of operation of the air conditioner 50. Vertical line B represents the start time of business for the company. Horizontal line C represents the target temperature. Line C is preferably set to a temperature in which the PMV value of occupants is in the range of -0.5 to +0.5. Line D represents the air-conditioned temperature actually discharged from the time the air conditioner 50 started operation until the start time of business. Line E represents a predetermined air-conditioned temperature. Line E is preferably an air-conditioned temperature that makes it easier to improve energy saving effects.

[0042] The measurement unit 104 is configured to measure the amount of energy consumed in conjunction with the operation of the air conditioner 50. The measurement unit 104 is also configured to estimate the amount of energy consumed by the air conditioner 50 at a time corresponding to the predicted first temperature (in this embodiment, 1 hour later) when the predicted first temperature is estimated.

[0043] The measured or estimated energy consumption is used to determine a control method that reduces energy consumption within the first or second range described later. The control unit 106 has a configuration that allows it to communicate control signals to the air conditioner 50 based on the prediction accuracy. Specifically, when the prediction accuracy is high, the control unit 106 controls the air conditioner so that the PMV value falls within a first range. More specifically, when the accuracy is high, the control unit 106 controls the air conditioner so that the air conditioning index is within the first range and the amount of energy consumed is as low as possible.

[0044] In this embodiment, if the prediction accuracy is low, the control unit 106 controls the air conditioning index so that it falls within the second range. Furthermore, if the prediction accuracy is high, it is preferable that the control unit 106 controls the air conditioning index so that it is within the second range and the amount of energy consumed is as low as possible.

[0045] In the case of low precision, the control unit 106 may change the control content of the air conditioner 50 to a predetermined setting for the operator. This control content may be one that improves comfort or one that improves energy saving. Furthermore, control that improves comfort or energy saving is preferable to control in the case of high precision.

[0046] Specifically, in the case of low accuracy, it is preferable that the control unit 106 changes the set temperature of the air conditioner 50 based on the air conditioning index. In this case, it is preferable that the set temperature improves the energy-saving effect of the air conditioner 50. Alternatively, in the case of low accuracy, it is preferable that the control unit 106 changes the control content of the air conditioner 50 to a predetermined setting for the operator.

[0047] The first range is a predetermined range of PMV values. In this embodiment, the first range is preferably the calculated PMV value in the range of -0.5 to +0.5. However, the PMV value may be other than the values ​​mentioned above.

[0048] The second range is a narrower range of PMV values ​​than the first range, with a PMV value of 0 as the median. The PMV value in the second range may be 0. Note that the PMV value may be any value other than those mentioned above.

[0049] Furthermore, when the control unit 106 controls the air conditioner 50 to fall within a first range or a second range, it can control the calculated PMV value to fall within a first range or a second range by changing at least one of the start time of the air conditioner 50, the stop time of the air conditioner 50, and the set temperature of the air conditioner 50.

[0050] The camera is installed to capture images of the interior of the building, with at least one camera, and is a device for capturing images of people inside the building. Furthermore, the camera is configured to transmit captured image information to the control device 100. In this embodiment, it is preferable that the camera is capable of capturing images of the interior of the building.

[0051] The camera may be an existing camera with camera functionality located within the building, or it may be incorporated into the control device 100, or it may be configured independently. In this embodiment, the case in which the camera is configured independently of the control device 100 will be described.

[0052] Furthermore, it is preferable that the camera be installed in the upper space so that it can overlook the room being filmed from above. Specifically, it is preferable that the camera be installed on the ceiling of the room being filmed. However, the camera may be installed in a location other than the ceiling of the room being filmed. In addition, a configuration other than a camera is also acceptable as long as it can capture the number of people in the room, the amount of clothing they are wearing, and their activity level.

[0053] The measuring instrument is a device configured to measure the indoor temperature. In addition to the indoor temperature, the measuring instrument may also measure other elements necessary for calculating the PMV value. Description of the action Next, the operation of the air conditioning system 10 with the above configuration will be described. First, the air conditioner 50 will be described, second, the camera will be described, third, the control of the control device 100 will be described, and fourth, the learning method of the learning model will be described.

[0054] First, the air conditioner 50 draws in indoor air from the office by rotating its air conditioning fan. The air drawn in includes air that has been heated by occupants and air that has been discharged from electronic devices installed in the office.

[0055] The inhaled air is cooled in the heat exchange unit (not shown). Specifically, the inhaled air loses heat to the refrigerant circulating between it and the outdoor unit (not shown), causing its temperature to drop. The refrigerant, having lost heat, releases heat into the outside air in the outdoor unit. The refrigerant that has released heat then again absorbs heat from the inhaled air in the heat exchange unit.

[0056] The air cooled in the heat exchange section is sent into the office by an air conditioner fan (not shown). The air sent into the office is conditioned to ensure a comfortable environment for occupants based on the PMV value. The air sent into the office is then drawn back into the air conditioner 50.

[0057] It is preferable that the air conditioner 50 is controlled so that the room temperature reaches a predetermined temperature by the time the company starts working. In other words, it is preferable that the air conditioner 50 is started before the company starts working.

[0058] Next, the camera control will be explained. When power is supplied to the camera, it continues to photograph the target space. If a person in the building is photographed, the camera transmits the image information of the person to the control device 100. In other words, each time a person in the building is photographed, the camera transmits the image information of the person to the control device 100.

[0059] Next, the control in the control device 100 will be explained with reference to Figure 3. When processing in the control device 100 begins, the acquisition unit 101 performs the process of acquiring image information, etc., at an arbitrary timing (S1). The image information, etc., may include one occupant or may include multiple occupants.

[0060] The image information acquired by the acquisition unit 101 is stored in the storage unit 102. Based on the stored image information, the calculation unit 107 performs a process to calculate the PMV value of the occupants (S2). The calculated PMV value may be the PMV value of one person or the average of multiple PMV values.

[0061] Once processing in S2 is complete, the prediction unit 103 estimates the first predicted temperature based on the stored image information, etc. (S3). Specifically, it performs a process to estimate the first predicted temperature using machine learning.

[0062] Once the first predicted temperature is estimated, the accuracy unit 105 performs a process to calculate the prediction accuracy of the first predicted temperature (S4). Specifically, it calculates the temperature difference between the first predicted temperature and the first temperature at the time corresponding to the first predicted temperature.

[0063] Furthermore, the accuracy unit 105 determines whether the calculated prediction accuracy value exceeds a predetermined threshold (S5). If the prediction accuracy value exceeds the threshold, the accuracy unit 105 determines that the accuracy is high. If the prediction accuracy value does not exceed the threshold, the accuracy unit 105 determines that the accuracy is low.

[0064] Once the prediction accuracy is calculated, the control unit 106 performs a process to control the air conditioner 50. Specifically, the process differs depending on whether the accuracy is high or low, so the process of the control unit 106 will be explained in each case.

[0065] First, the processing of the control unit 106 in the case of high accuracy will be explained. The measurement unit 104 determines the control content of the air conditioner 50 that has the lowest energy consumption while the PMV value is within the first range (S6).

[0066] Based on the determined control content, the control unit 106 controls the air conditioner 50 (S7). In other words, the control unit 106 performs control such that the PMV value is within the first range and the energy consumption of the air conditioner 50 is as low as possible.

[0067] Next, the processing of the control unit 106 in the case of low accuracy will be described. The measurement unit 104 determines the control content of the air conditioner 50 that has the lowest energy consumption while the PMV value is within the second range (S6). Based on the determined control content, the control unit 106 controls the air conditioner 50 (S7). In other words, it is preferable that the control unit 106 transmits a signal indicating the control content for the air conditioner 50 in which the PMV value is within the second range and the amount of energy consumed is low.

[0068] Next, we will discuss machine learning in the context of learning models. In this embodiment, the explanation applies to an example where machine learning of a learning model is performed in an information processing device different from the control device 100. The learning model that has undergone machine learning is stored in the storage unit 102 before processing by the control device 100.

[0069] Furthermore, a model that has undergone further machine learning after control by the control device 100 may be stored in the memory unit 102. In this case, the previously stored learning model is replaced by the learning model that has undergone further machine learning.

[0070] Furthermore, the machine learning of the learning model may be performed in different information processing devices as described above, or it may be performed in the control device 100. If machine learning is performed in the control device 100, the control device 100 will be equipped with a machine learning unit for performing machine learning. Alternatively, machine learning may be performed on one of the learning models in different information processing devices, and machine learning may be performed on the other in the control device 100.

[0071] In machine learning models, training data is used to determine if a person is the same person before and after they move. Specifically, training data may be used that determines the distance a person moves based on the distance estimated from preceding and succeeding frames of an image, or training data may be used that determines the same person based on similarity.

[0072] Regarding the specific machine learning methods used in the learning model, known supervised learning methods may be employed, and there are no limitations on the specific content of the computational processing in supervised learning. Similarly, regarding the method for creating the training data for the learning model, known methods may be used, and there are no limitations on the specific method of creation.

[0073] Description of effects The air conditioning system 10 with the above configuration allows for different control in high-precision and low-precision settings. By performing different control in high-precision and low-precision settings, the air conditioning system 10 can easily improve comfort. Furthermore, by performing different control in high-precision and low-precision settings, the air conditioning system 10 can easily reduce energy consumption. In other words, by performing different control in high-precision and low-precision settings, the air conditioning system 10 can easily achieve both improved comfort and reduced energy consumption.

[0074] Furthermore, by using machine learning to estimate the first predicted temperature, the prediction unit 103 can estimate the first predicted temperature with higher accuracy compared to estimating the first predicted temperature without using machine learning. Moreover, by using machine learning to estimate the first predicted temperature, it is easier to reduce energy consumption. In other words, by using machine learning to estimate the first predicted temperature, the air conditioning system 10 can easily achieve both improved comfort and reduced energy consumption.

[0075] Furthermore, the control unit 106 can change the start time of the air conditioning unit 10, the stop time of the air conditioning unit 10, and the set temperature of the air conditioning unit 10 (hereinafter also referred to as variables, etc.). By changing the variables, etc., it is possible to increase the number of options for air conditioning indicators. In other words, it is easier to achieve both improved comfort and reduced energy consumption.

[0076] Furthermore, by changing the start time of the air conditioning unit 10, the stop time of the air conditioning unit 10, and the set temperature of the air conditioning unit 10, it is possible to achieve control within the first range that maximizes both comfort and energy saving.

[0077] Furthermore, by changing the start time of the air conditioning unit 10, the stop time of the air conditioning unit 10, and the set temperature of the air conditioning unit 10, a variety of control methods can be implemented within the second range, while maintaining both comfort and energy-saving effects, such as control prioritizing comfort or control prioritizing energy-saving effects.

[0078] It should be noted that the technical scope of the present invention is not limited to the embodiments described above, and various modifications can be made without departing from the spirit of the invention. For example, the present invention is not limited to those applied to the embodiments described above, but may also be applied to embodiments that combine these embodiments as appropriate, and is not particularly limited. [Explanation of symbols]

[0079] 10...Air conditioning unit, 50...Air conditioner, 100...Control device, 101...Acquisition unit, 102...Storage unit, 103...Prediction unit, 104...Measurement unit, 105...Accuracy unit, 106...Control unit, 107...Calculation unit.

Claims

1. An air conditioning device that adjusts a first temperature, which is the temperature of the target space, A prediction unit that determines a predicted first temperature, which is the first temperature after a predetermined time, based on the first temperature, a second temperature which is the temperature outside the target space, the set temperature of the air conditioning unit, and the number of people staying in the target space, A precision unit that determines the prediction accuracy based on the predicted first temperature and the first temperature after a predetermined time, A control unit that controls the air conditioning system so that the air conditioning index falls within a predetermined first range, If the prediction accuracy exceeds a predetermined threshold, control is performed so that the air conditioning index falls within the first range. If the prediction accuracy does not exceed the threshold, the control unit controls the air conditioning index to fall within a second range that is narrower than the first range, or controls the air conditioning index without using it. An air conditioning system characterized by being provided with [a specific feature].

2. The air conditioning device according to claim 1, characterized in that the prediction unit estimates the predicted first temperature by inputting the first temperature, the second temperature, the set temperature of the air conditioning device, and the number of people staying in the target space into a trained learning model that has been trained by machine learning to estimate the predicted first temperature after a predetermined time in the target space.

3. The air conditioning system according to claim 1, characterized in that the control unit controls the air conditioning index to fall within the first range or the second range by changing at least one of the start time of the air conditioning system, the stop time of the air conditioning system, and the set temperature of the air conditioning system.

4. An air conditioning device that adjusts a first temperature, which is the temperature of the target space, A prediction unit that determines a predicted first temperature, which is the first temperature after a predetermined time, based on the first temperature, a second temperature which is the temperature outside the target space, the set temperature of the air conditioning unit, and the number of people staying in the target space, A precision unit that determines the prediction accuracy based on the predicted first temperature and the first temperature after a predetermined time, A control unit that controls the air conditioning system so that the air conditioning index falls within a predetermined first range, If the prediction accuracy exceeds a predetermined threshold, control is performed so that the air conditioning index falls within the first range. The control unit controls the set temperature of the air conditioning unit if the prediction accuracy does not exceed the threshold, An air conditioning system characterized by being provided with [a specific feature].

Images