Air conditioning equipment

Abstract

The invention discloses air conditioning equipment which comprises a sensing module, a first characteristic parameter acquisition module, a second characteristic parameter acquisition module, a target parameter acquisition module and a prediction module. Wherein the first characteristic parameter acquisition module is configured to acquire a dressing grade, a somatosensory wind speed and an indoor environment background radiation temperature of a tested human body target based on a thermal image; the second characteristic parameter acquisition module is configured to acquire the environment temperature and the environment humidity of the indoor environment; the target parameter acquisition module is configured to acquire combinations of target sensible temperature and sensible humidity of a tested human body, and an integer index is used for representing a target label of each combination; the prediction module is provided with a prediction model generated based on the dressing grade, the somatosensory wind speed, the background radiation temperature, the environment temperature, the environment humidity and the target label. The integer index corresponding to the maximum probability value output by the prediction model is used for outputting the human body target comfort degree category. When a small sample environment is dealt with, the air conditioning equipment provided by the invention can effectively reduce the over-fitting risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air conditioning technology, in particular to an air conditioning equipment. BACKGROUND

[0002] The PWM model (Predicted Mean Vote Model) is a thermal comfort prediction model based on human body heat balance theory. It predicts the average thermal sensation of a group in a specific thermal environment by considering parameters such as air temperature, relative humidity, mean radiant temperature, air flow rate, human metabolic rate, and clothing thermal resistance, and quantifies it into multiple grades. It is widely used in thermal environment evaluation and air conditioning system design. Combining the PWM model with machine learning can take full advantage of machine learning in data processing and pattern recognition, improve or extend the PWM model, and achieve more personalized and dynamic thermal environment control.

[0003] Chinese patent application CN118035932A discloses an individual comfort prediction method. By obtaining and preprocessing individual thermal comfort related data, a data set is constructed. Using ensemble learning technology, combining naive Bayes, random forest, support vector machine as base learner, and limit number as meta-learner, a prediction model is constructed. The prediction model further improves the accuracy and generalization ability of predicting individual comfort in a specific thermal environment through training and cross-validation.

[0004] The above-mentioned patent discloses a method that integrates multiple sensors and multiple models. However, for limited use scenarios of air conditioning products, the number of samples for training the model is usually in the order of tens, hundreds, or thousands. When the sample size is too complex or the data feature dimension is very high, the risk of overfitting is very high, making the model too sensitive to noise or outliers in the training data. In particular, after being configured in the actual use environment, it cannot respond to new conditions, further leading to the failure of the control strategy corresponding to the prediction result in actual application.

[0005] The above information disclosed in the background of the application is only used to increase the understanding of the background of the application, and therefore, it can include prior art known to those of ordinary skill in the art. SUMMARY

[0006] In view of the problem that the risk of overfitting is very high for complex models trained based on small-scale training samples, and after being configured in the actual use environment, it cannot respond to new conditions, leading to the failure of the control strategy corresponding to the prediction result in actual application, an air conditioning equipment is designed and provided.

[0007] To achieve the above-mentioned application purposes, the present application adopts the following technical solutions:

[0008] The air conditioning equipment comprises a sensing module. The sensing module comprises an infrared temperature imaging sensor configured to form a thermal image of an indoor environment.

[0009] In one or more embodiments of the present application, the air conditioning equipment further comprises a first feature parameter acquisition module, a second feature parameter acquisition module, a target parameter acquisition module, and a prediction module; wherein the first feature parameter acquisition module is configured to obtain a clothing level of the test human target, a felt wind speed, and a background radiation temperature of the indoor environment based on the thermal image; the second feature parameter acquisition module is configured to obtain an ambient temperature and an ambient humidity of the indoor environment; the target parameter acquisition module is configured to obtain a combination of the felt temperature and the felt humidity of the test human target, and to represent a target label of each combination with an integer index; and the prediction module has a prediction model generated based on the clothing level, the felt wind speed, the background radiation temperature, the ambient temperature, the ambient humidity, and the target label. The integer index corresponding to the maximum probability value output by the prediction model is the output human target comfort level.

[0010] In one or more embodiments of the present application, the first feature parameter acquisition module is configured to identify and select the test human target in the thermal image of the indoor environment, and to determine the felt wind speed of the test human target corresponding to the air supply mode according to the air supply mode.

[0011] In one or more embodiments of the present application, the first feature parameter acquisition module is configured to perform the following steps: if the air supply mode is the wind-avoiding-human-blowing mode, the felt wind speed is a preset minimum felt wind speed; if the air supply mode is a non-wind-avoiding-human-blowing mode, the relative distance and the relative angle between the test human target and the air outlet are identified; and the felt wind speed is estimated based on the relative distance, the relative angle, and the real-time air supply speed.

[0012] In one or more embodiments of the present application, the first feature parameter acquisition module is configured to determine a target temperature of each pixel point in the human infrared thermal image, to screen a maximum pixel temperature value in the target temperature and to calculate an average pixel temperature value of the human infrared thermal image; and when a proportional relationship between the average pixel temperature value and the maximum pixel temperature value satisfies one of a plurality of preset clothing amount ratio threshold conditions, the clothing amount of the test human target is determined as a clothing level corresponding to the clothing amount ratio threshold condition.

[0013] In one or more embodiments of the present application, the first feature parameter acquisition module is configured to filter out the human infrared thermal image in the thermal image of the indoor environment, and to generate the background radiation temperature based on the thermal image of the indoor environment after filtering out the human infrared thermal image.

[0014] In one or more embodiments of the present application, the second feature parameter acquisition module classifies the ambient temperature based on a preset temperature level, and outputs the ambient temperature level.

[0015] In one or more embodiments of the present application, the second feature parameter acquisition module classifies the ambient humidity based on the preset humidity level, and outputs the ambient humidity level.

[0016] In one or more embodiments of the present application, the target parameter acquisition module is configured to: acquire a thermal comfort level corresponding to the target thermal sensation temperature of the test human body; acquire a humidity comfort level corresponding to the target humidity sensation temperature of the test human body; and assign a corresponding integer index to each combination of the thermal comfort level and the humidity comfort level to represent the target label of each combination.

[0017] In one or more embodiments of the present application, the target parameter acquisition module is configured to: represent the target thermal comfort level of the test human body by a thermal comfort level vector, wherein the value of one bit in the thermal comfort level vector is 1 and the values of the remaining bits are 0; represent the target humidity comfort level of the test human body by a humidity comfort level vector, wherein the value of one bit in the humidity comfort level vector is 1 and the values of the remaining bits are 0; and convert each combination of the thermal comfort level vector and the humidity comfort level vector into an integer index, wherein the value of the integer index represents the target label of each combination.

[0018] In one or more embodiments of the present application, the air conditioning device further comprises a control module. The control module is configured to output a control parameter based on the user comfort level category and the air supply mode output by the prediction model.

[0019] In one or more embodiments of the present application, the control parameter comprises at least one of a target temperature, a target humidity, a set air speed, a set air direction, a condensing temperature, an evaporating temperature, a throttling element opening degree, and a compressor rotation speed.

[0020] In one or more embodiments of the present application, the control module is further configured to acquire an integer index corresponding to the maximum probability value and an integer index corresponding to the second maximum probability value output by the prediction model, and calculate the difference between the maximum probability value and the second maximum probability value, and when the difference is lower than a preset difference threshold, suspend the output of the control parameter for at least one determination period and keep the current operating parameter unchanged.

[0021] Compared with the prior art, the air conditioning equipment provided by the application has the advantages and positive effects that: the first feature parameter acquisition module of the air conditioning equipment provided by the application acquires the target feature parameter of the human body based on the indoor environment thermal image, the second feature parameter acquisition module extracts the environment temperature and the environment humidity, and further describes the indoor environment state; this multi-source data acquisition mode can effectively increase the data diversity and reduce the risk of overfitting in the data acquisition stage; the target parameter acquisition module uses an integer index to represent the target label of each combination, so as to more accurately quantify the comfort level of the user, make the prediction model better learn the relationship between the user comfort level and the environment factors in the training process, and improve the generalization ability of the model; when dealing with a small sample environment, the air conditioning equipment provided by the application can effectively reduce the risk of overfitting.

[0022] Other features and advantages of the present application will become more apparent from the following detailed description of some embodiments of the present application, taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings described below are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative labor based on these drawings.

[0024] Figure 1 The structural schematic block diagram of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0025] Figure 2 The structural schematic diagram of the sensing module in the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0026] Figure 3 The flowchart of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0027] Figure 4 The flowchart of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0028] Figure 5 The flowchart of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0029] Figure 6 The flowchart of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0030] Figure 7 The flowchart of the air conditioning equipment provided by some embodiments of the present application is shown in the figure.

[0031] Figure 8An example of data of the first feature parameter acquisition module, the second feature parameter acquisition module, the target parameter acquisition module and the prediction module in the air conditioning device provided by some embodiments of the present application in combination with the air supply configuration;

[0032] Figure 9 An example of converting the combination of the individual thermal comfort level vector and the thermal humidity comfort level vector into an integer index in the air conditioning device provided by some embodiments of the present application;

[0033] Figure 10 A flowchart of the air conditioning device provided by some embodiments of the present application;

[0034] In the figure:

[0035] 10, air conditioning device; 101, sensing module; 102, first feature parameter acquisition module; 103, second feature parameter acquisition module; 104, target parameter acquisition module; 105, prediction module; 111, infrared temperature measurement imaging sensor. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0037] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0038] The terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, "a plurality of" means two or more.

[0039] In the description of the application, it is necessary to point out that, unless otherwise explicitly specified and limited, the terms "mount", "connect", "connection" should be understood broadly, for example, it can be fixed connection, or detachable connection, or integrally connected; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through intermediate medium, or the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0040] In the present application, unless otherwise explicitly specified and limited, the "on" or "under" of the first feature to the second feature can include that the first and second features are in direct contact, or that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, the "on", "above" and "over" of the first feature to the second feature includes that the first feature is directly above and obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. The "under", "below" and "under" of the first feature to the second feature includes that the first feature is directly below and obliquely below the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature.

[0041] The following disclosure provides many different embodiments or examples for implementing different structures of the application. For simplicity of the present disclosure, the components and settings of particular examples are described in the following. Of course, they are only examples and the purpose is not to limit the application. In addition, the present application can repeatedly refer to numbers and / or letters in different examples. Such repetition is for the purpose of simplification and clarity, and in itself does not indicate the relationship between the various embodiments and / or settings discussed. In addition, the present application provides examples of various specific processes and materials, but those skilled in the art can realize the application of other processes and / or the use of other materials.

[0042] For a complex model trained based on a small-scale training sample, the risk of overfitting is very high. After being configured in the actual use environment, it cannot respond to new conditions, resulting in the problem that the control strategy corresponding to the prediction result fails in actual application. An air conditioning device is designed and provided.

[0043] The air conditioning device 10 is a system that performs a refrigeration cycle of the air conditioning device 10 by using a compressor, a condenser, a throttling device, and an evaporator. The refrigeration cycle includes a series of processes involving compression, condensation, expansion, and evaporation to cool or heat an indoor space.

[0044] From the principle point of view, the low-temperature and low-pressure refrigerant enters the compressor, the compressor compresses the refrigerant gas into a high-temperature and high-pressure state and discharges the compressed refrigerant gas. The discharged refrigerant gas flows into the condenser. The condenser condenses the compressed refrigerant into a liquid phase, and the heat is released to the surrounding environment through the condensation process.

[0045] The throttling device expands the high-temperature and high-pressure liquid-phase refrigerant condensed in the condenser into a low-pressure liquid-phase refrigerant. The evaporator evaporates the refrigerant expanded in the throttling device and returns the refrigerant gas in a low-temperature and low-pressure state to the compressor. The evaporator can achieve a refrigeration effect by exchanging heat with a material to be cooled using latent heat of evaporation of the refrigerant. Throughout the cycle, the air conditioning apparatus 10 can adjust the temperature of an indoor space.

[0046] The outdoor unit of the air conditioning apparatus 10 refers to a portion of the refrigeration cycle including the compressor and the outdoor heat exchanger, and the indoor unit of the air conditioning apparatus 10 is provided in an air conditioning room, includes the indoor heat exchanger, and the throttling device can be provided in the indoor unit and the outdoor unit.

[0047] The indoor heat exchanger and the outdoor heat exchanger function as a condenser or an evaporator. When the indoor heat exchanger functions as a condenser, the air conditioning apparatus 10 functions as a heater in a heating mode, and when the indoor heat exchanger functions as an evaporator, the air conditioning apparatus 10 functions as a cooler in a cooling mode.

[0048] In an alternative embodiment, one or more compressors can be provided in each outdoor unit, and the compressor in operation is supplied with alternating current through a frequency conversion device. When the output frequency of the frequency conversion device changes, the rotational speed of the compressor changes, and different air conditioning capacities are achieved.

[0049] The outdoor unit further includes an outdoor fan and a four-way valve. In addition, a gas-liquid separator, a capillary tube, an oil separator, and other conventional components can be provided. The gas-liquid separator is a shell-shaped component for separating refrigerant into gas and liquid, and is usually provided at the suction side of the compressor. The outdoor heat exchanger is configured as a heat exchanger that exchanges heat between refrigerant flowing through the inside of the heat exchanger pipe and air (or other medium) guided by the outdoor fan. The outdoor fan can be a form of an axial indoor fan, a cross-flow indoor fan, or other alternative indoor fan, and is usually provided near the outdoor heat exchanger.

[0050] The four-way valve is a valve that switches the flow direction of the refrigerant according to the operation mode of the air conditioning apparatus 10, i.e., in the cooling mode, the discharge side of the compressor is connected to one end of the outdoor heat exchanger through the four-way valve and the like, and the suction side of the compressor is connected to one end of the indoor heat exchanger through the four-way valve and the like, whereby the outdoor heat exchanger functions as a condenser and the indoor heat exchanger functions as an evaporator.

[0051] Similarly, in the heating mode, the discharge side of the compressor is connected to one end of the indoor heat exchanger through the four-way valve and the like, and the suction side of the compressor is connected to one side of the outdoor heat exchanger through the four-way valve and the like, whereby the indoor heat exchanger functions as a condenser and the outdoor heat exchanger functions as an evaporator.

[0052] The indoor electronic expansion valve and the outdoor electronic expansion valve are valves that depressurize the refrigerant flowing into the valve body itself, and are provided in a pipe through which a liquid refrigerant or a gas-liquid two-phase refrigerant flows.

[0053] The oil separator is used to separate the lubricating oil in the refrigerant discharged from the compressor, and is usually provided on the discharge side of the compressor. The lubricating oil separated by the oil separator can be guided to the gas-liquid separator through a pipeline. A one-way valve can also be provided to guide the separated refrigerant to the four-way valve.

[0054] The outdoor control circuit is usually provided in an electrical box with good sealing performance. The outdoor control circuit includes a processor, a storage unit, an input / output interface, a communication interface, and the like. The processor can be a special-purpose processor, a central processing unit (CPU), or the like. The processor can access the storage unit to execute instructions or application programs stored in the storage unit to implement related functions. The storage unit can include a volatile memory and / or a non-volatile memory. The input / output interface can be in communication connection with various sensors provided in the outdoor unit to receive detection values of various sensors provided in the outdoor unit.

[0055] In an optional embodiment, the indoor unit is matched with a wire controller, which is fixedly installed on a wall of the air-conditioned room. The wire controller is provided with an operation interface for inputting a set temperature and an operation mode, and a display interface for displaying a real-time temperature of the air-conditioned room and a running state of the air conditioning device 10.

[0056] In an optional embodiment, the indoor unit is matched with a remote controller, which is in communication connection with the indoor unit. The remote controller is provided with a key for inputting a set temperature and an operation mode, and a display interface for displaying a real-time temperature of the air-conditioned room and a running state of the air conditioning device 10.

[0057] In an optional embodiment, the indoor unit is matched with a mobile control terminal, which is in communication connection with the indoor unit. The mobile control terminal has an application interface, through which a set temperature and an operation mode can be input, and a real-time temperature of the air-conditioned room or a running state can be displayed.

[0058] In an optional embodiment, the mobile control terminal can be a computer, a tablet computer, a smart phone, a wearable device, or the like.

[0059] One or more of the wire controller, the remote controller, and the mobile control terminal are provided with a temperature sensor and a humidity sensor to detect an ambient temperature and an ambient humidity.

[0060] The indoor unit is provided with an indoor unit control circuit, and the indoor unit control circuit is preferably provided with an indoor controller. The indoor controller is configured to drive the indoor fan to work, display various parameters on the display panel, realize human-computer interaction, receive and process sampling signals of various sensors, and realize necessary communication functions.

[0061] The indoor unit control circuit also includes storage units, processors, input / output interfaces, communication interfaces, and other electrical components.

[0062] The storage unit can include volatile memory and / or non-volatile memory. The storage unit is configured to store instructions or data associated with at least one component of the indoor unit, such as storing an application program. For example, the application program can be used to adjust the temperature of the air-conditioned room through different speed gears of the indoor fan.

[0063] The indoor processor can be a dedicated processor, a central processing unit (CPU), or the like. The indoor processor can access the storage unit to execute instructions stored in the storage unit to achieve related functions.

[0064] As shown in Figure 1 In one or more embodiments of the present application, the air conditioning equipment 10 includes a sensing module 101. As shown in Figure 2 The sensing module 101 includes an infrared temperature imaging sensor 111, which is based on the principle of infrared radiation. The infrared sensor detects the infrared radiation emitted by the indoor (air-conditioned room) object, converts it into an electrical signal through the optical system and detector in the infrared temperature imaging sensor 111, and forms a thermal image.

[0065] In one or more embodiments of the present application, the sensing module 101 is a combination of one or more of the infrared temperature imaging sensor 111, the array thermoelectric sensor, and the thermistor sensor. The number of each of the infrared temperature imaging sensor 111, the array thermoelectric sensor, and the thermistor sensor is not limited here.

[0066] In one or more embodiments of the present application, the air conditioning equipment 10 further includes a first feature parameter acquisition module 102, a second feature parameter acquisition module 103, a target parameter acquisition module 104, and a prediction module 105.

[0067] In one or more embodiments of the present application, the first feature parameter acquisition module 102 is configured to obtain the clothing level of the test human target, the body feeling wind speed, and the indoor environment background radiation temperature based on the thermal image.

[0068] In one or more embodiments of the present application, the second feature parameter acquisition module 103 is configured to obtain the ambient temperature and the ambient humidity of the indoor environment.

[0069] In one or more embodiments of the present application, the target parameter acquisition module 104 is configured to acquire a combination of the target body sensible temperature and the target body sensible humidity, and to represent the target label of each combination by an integer index.

[0070] In one or more embodiments of the present application, the prediction module 105 has a prediction model generated based on the clothing level, the sensible wind speed, the background radiation temperature, the environment temperature, the environment humidity, and the target label, and the integer index corresponding to the maximum probability value output by the prediction model is the output human body target comfort class.

[0071] The first feature parameter acquisition module 102 of the air conditioning equipment 10 provided in the present application acquires the target feature parameter of the test human body based on the indoor environment thermal image, and the second feature parameter acquisition module 103 extracts the environment temperature and the environment humidity to further describe the indoor environment state; such a multi-source data acquisition mode can effectively increase the data diversity and reduce the risk of overfitting in the data acquisition stage; the target parameter acquisition module 104 represents the target label of each combination by an integer index, thereby more accurately quantifying the comfort class of the user, enabling the prediction model to better learn the relationship between the user comfort and the environmental factors in the training process, and improving the generalization ability of the model; when dealing with a small sample environment, the air conditioning equipment 10 provided in the present application can effectively reduce the risk of overfitting.

[0072] In one or more embodiments of the present application, the first feature parameter acquisition module 102 is configured to identify and select the test human body target in the indoor environment thermal image, and to determine the sensible wind speed corresponding to the test human body target and the air supply mode according to the air supply mode.

[0073] In one or more embodiments of the present application, the first feature parameter acquisition module 102 can identify and select the test human body target in the indoor environment thermal image through image processing.

[0074] For example, the first feature parameter acquisition module 102 pre-processes the indoor environment thermal image, removes image noise using a filter, and improves the visibility of the test human body target in the indoor environment thermal image using a contrast enhancement algorithm. According to the infrared radiation characteristics of the test human body target, the test human body target is usually higher in temperature than the background object, a temperature threshold is set, the image is binarized, and the possible heat source area is separated out. Small noise points in the heat source area are removed using expansion, corrosion, and other operations, the holes in the heat source area are filled, the connected component analysis is performed on the binary image of the heat source area, the continuous pixel area is identified, and each area corresponds to a potential test human body target. The features of each connected area are extracted to confirm the test human body target.

[0075] Exemplarily, the test human target can be identified based on the features of the connected region, including area, shape and temperature distribution, by a convolutional neural network.

[0076] In one or more embodiments of the present application, the first feature parameter acquisition module 102 performs a plurality of steps as shown in Figure 3 to determine the body sensation wind speed corresponding to the test human target and the air supply mode.

[0077] Step S101: Determine whether the air supply mode is the wind-avoiding-human-blowing mode.

[0078] Step S102: If the air supply mode is the wind-avoiding-human-blowing mode, the body sensation wind speed is a preset minimum body sensation wind speed.

[0079] Step S103: If the air supply mode is the non-wind-avoiding-human-blowing mode, the relative distance and the relative angle between the test human target and the air supply port are identified, and the body sensation wind speed is determined based on the relative distance, the relative angle and the real-time wind speed, as shown in step S104.

[0080] In one or more embodiments of the present application, the minimum body sensation wind speed can be set to 0.3 m / s.

[0081] In one or more embodiments of the present application, based on the camera calibration information of the infrared temperature measurement imaging sensor 111, the image coordinate system is converted into the actual coordinate system, and the centroid of the test human target region is taken as the relative position of the test human target. Further, the relative distance and the relative angle between the test human target and the air supply port can be determined.

[0082] In one or more embodiments of the present application, when determining the body sensation wind speed, it is assumed that the air flow in the air conditioning room is uniformly diffused. Under the condition of uniform diffusion of air flow, it is approximately considered that the total energy of air is conserved during the diffusion process, and the air flow output by the air supply port of the indoor unit after adjustment is in spherical or conical diffusion distribution. The energy of air flow is proportional to the square of the wind speed, and the diffusion area of air flow is proportional to the square of the distance. Based on the above assumptions, in the axial direction of the air supply port of the indoor unit of the air conditioning device 10, it can be determined that the body sensation wind speed at a certain position and the square of the relative distance between the test human target and the air supply port are inversely proportional.

[0083] Exemplarily, the relative distance between the air supply port of the indoor unit of the air conditioning device 10 and the test human target increases by 1 times, and the body sensation wind speed on the axis of the air supply port attenuates to 1 / 4 of the wind speed of the indoor unit.

[0084] In the direction of the air outlet axis, the estimated body sensation wind speed is maximum. As the angle between the test human target and the air outlet increases, the body sensation wind speed gradually decreases. Based on the above assumption, the first characteristic parameter acquisition module 102 is configured with a correction coefficient corresponding to the relative angle. The product of the correction coefficient and the relative angle is the estimated body sensation wind speed. The correction coefficient can be generated based on the cosine function value of the relative angle.

[0085] In one or more embodiments of the present application, when estimating the body sensation wind speed, the first characteristic parameter acquisition module 102 can also be configured to estimate whether there is an obstacle between the air outlet and the test human target. If there is no obstacle, the body sensation wind speed is estimated in the above manner. If there is an obstacle, the estimated body sensation wind speed is corrected according to the area of the air outlet blocked by the obstacle to call a pre-set obstacle correction coefficient. The larger the area of the air outlet blocked by the obstacle, the larger or more dense the obstacle, and the larger the correction amount of the obstacle correction coefficient, the smaller the estimated body sensation wind speed.

[0086] As Figure 4 As shown in steps S201 to S205, in one or more embodiments of the present application, the first characteristic parameter acquisition module 102 is configured to generate a human infrared thermal image, determine the target temperature of each pixel point in the human infrared thermal image, screen the maximum pixel temperature value in the target temperature, and calculate the average pixel temperature value of the human infrared thermal image; when the proportional relationship between the average pixel temperature value and the maximum pixel temperature value satisfies one of a plurality of pre-set clothing amount ratio threshold conditions, it is confirmed that the clothing amount of the test human target is the clothing level corresponding to the clothing amount ratio threshold condition.

[0087] In one or more embodiments of the present application, the clothing level is divided into four levels, i.e. the first clothing level, the second clothing level, the third clothing level and the fourth clothing level. The first clothing level corresponds to short-sleeved T-shirts, shorts and other exposed limbs; the second clothing level corresponds to shirts, trousers and other single-layer clothing that cover the limbs; the third clothing level corresponds to long-sleeved, long-pants and other double-layer clothing that cover the limbs; and the fourth clothing level corresponds to down jackets, sweaters, wool pants and other multi-layer clothing that cover the limbs. In a scenario where control accuracy is required to be higher, more clothing levels can also be set.

[0088] After selecting the maximum pixel temperature value from the target temperature range and calculating the average pixel temperature value of the human infrared thermal image, the system first checks if the ratio of the average pixel temperature value to the maximum pixel temperature value is greater than a first clothing level threshold. If it is, the clothing level of the test human target is confirmed as the first clothing level. If it is not greater than the first clothing level threshold, the system checks if the ratio of the average pixel temperature value to the maximum pixel temperature value is greater than a second clothing level threshold. If it is, the clothing level of the test human target is confirmed as the second clothing level. If it is not greater than the second clothing level threshold, the system checks if the ratio of the average pixel temperature value to the maximum pixel temperature value is greater than a third clothing level threshold. If it is, the clothing level of the test human target is confirmed as the third clothing level; otherwise, it is the fourth clothing level.

[0089] In one or more embodiments of this application, the first feature parameter acquisition module 102 is configured to filter out human infrared thermal images from an indoor environmental thermal image and generate a background radiation temperature based on the indoor environmental thermal image after filtering out the human infrared thermal images. Background radiation temperature is the average temperature of infrared radiation emitted by objects in the indoor environment, reflecting the overall radiation level of the indoor environment, and is related to room temperature, object surface temperature, and radiation characteristics. Generally speaking, the higher the indoor temperature, the higher the background radiation temperature. In particular, heaters, lights, and stoves can increase the indoor background radiation temperature, and ventilation and insulation measures can also affect the indoor temperature distribution, thereby affecting the background radiation temperature.

[0090] like Figure 5 Steps S301 to S305, and Figure 6 As shown in steps S306 to S310, in one or more embodiments of this application, the background radiation temperature is converted using the actual ambient temperature, and further classified. For example, when the background radiation temperature is lower than the indoor ambient temperature, and the temperature difference between the background radiation temperature and the indoor ambient temperature exceeds a set range, the background radiation temperature is classified as a first background radiation temperature level; when the temperature difference between the background radiation temperature and the indoor ambient temperature does not exceed a set range, the background radiation temperature is classified as a second background radiation temperature level; and when the background radiation temperature is higher than the indoor ambient temperature, and the temperature difference between the background radiation temperature and the indoor ambient temperature exceeds a set range, the background radiation temperature is classified as a third background radiation temperature level. By calculating the temperature difference, the converted data can highlight the inherent variation pattern of the background radiation temperature, making it more structured.

[0091] In one or more embodiments of this application, the second feature parameter acquisition module 103 classifies the ambient temperature based on a preset temperature level and outputs the ambient temperature level.

[0092] Exemplarily, the ambient temperature can be classified into 5 levels. For example, below 18 degrees Celsius is the first ambient temperature level, 18-23 degrees Celsius is the second ambient temperature level, 23-26 degrees Celsius is the third temperature level, 26 to 31 degrees Celsius is the fourth temperature level, and above 31 degrees Celsius is the fifth temperature level.

[0093] The levels of the ambient temperature can be adjusted according to actual use conditions. When the ambient temperature is detected, the level of the ambient temperature can be confirmed.

[0094] In one or more embodiments of the present application, the second feature parameter acquisition module 103 classifies the ambient humidity based on the preset humidity levels and outputs the ambient humidity level.

[0095] Exemplarily, the ambient humidity can be classified into 5 levels. For example, below 20% RH is the first ambient humidity level, 20-40% RH is the second ambient humidity level, 40-60% RH is the third ambient humidity level, 60-80% RH is the fourth ambient humidity level, and above 80% RH is the fifth ambient humidity level.

[0096] The levels of the ambient humidity can be adjusted according to actual use conditions. When the ambient humidity is detected, the level of the ambient humidity can be confirmed.

[0097] In one or more embodiments of the present application, the target parameter acquisition module 104 is configured to perform a plurality of steps as shown in Figure 7

[0098] Step S401: Obtain the level of comfort of the target thermal sensation temperature corresponding to the test human body.

[0099] Step S402: Obtain the level of comfort of the target thermal sensation humidity corresponding to the test human body.

[0100] Step S403: Assign a corresponding integer index to each combination of the level of comfort of the thermal sensation temperature and the level of comfort of the thermal sensation humidity to represent the target label of each combination.

[0101] A representative subject group is selected as the test human target, and different temperature and humidity combinations are simulated in the laboratory. The test human target evaluates the current thermal sensation temperature and thermal sensation humidity according to the subjective comfort level, and the subjective comfort level evaluation results of the subjects are recorded as target data samples.

[0102] An example of the data of the first feature parameter acquisition module 102, the second feature parameter acquisition module 103, the target parameter acquisition module 104, and the prediction module 105 in combination with the air supply configuration is shown in FIG. 8.

[0103] ​A unique integer index is assigned to each combination of the thermal sensation comfort level and the thermal humidity sensation comfort level as the target label of the combination. For example, a two-dimensional array can be used to store these mapping relationships.

[0104] In one or more embodiments of the present application, the target parameter obtaining module 104 is configured to represent the thermal sensation comfort level of the test human target in a thermal sensation comfort level vector. The value of one bit in the thermal sensation comfort level vector is 1, and the values of the other bits are 0. Similarly, the thermal humidity sensation comfort level of the test human target is represented in a thermal humidity sensation comfort level vector. The value of one bit in the thermal humidity sensation comfort level vector is 1, and the values of the other bits are 0. Each combination of the thermal sensation comfort level vector and the thermal humidity sensation comfort level vector is converted into an integer index, and the value of the integer index represents the target label of each combination. For example, as shown in Figure 9

[0105] The prediction module 105 has a prediction model generated based on the clothing level, the thermal wind speed, the background radiation temperature, the environmental temperature, the environmental humidity, and the target label.

[0106] In one or more embodiments of the present application, the prediction model is a neural network model. The training process of the prediction model includes a plurality of steps as shown in Figure 10

[0107] Step S501: Establish a data set. The data set contains feature parameters and target parameters. The feature parameters are the clothing level, the thermal wind speed, and the indoor environment background radiation temperature of the test human target obtained by the first feature parameter obtaining module 102, and the environmental temperature and the environmental humidity of the indoor environment obtained by the second feature parameter obtaining module 103. The target parameters are the combinations of the thermal sensation and the thermal humidity of the test human target obtained by the target parameter obtaining module 104.

[0108] The clothing level of the test human target is saved in the "cloth" column of the data sample, and the thermal wind speed, the indoor environment background radiation temperature, the environmental temperature, and the environmental humidity are saved in the "air_flow_speed", "background_temp", "env_temp", and "env_humid" columns of the data sample, respectively.

[0109] Each combination of the thermal sensation comfort level vector and the thermal humidity sensation comfort level vector of the test human target is converted into an integer index, and the value of the integer index represents the target label of each combination. For example, as shown in the figure, there are 9 categories.

[0110] ​​Step S502: The data set is divided into a training set and a test set according to a set proportion. The training set is used for model training to adjust model parameters and weights. The test set is used to verify the prediction effect of the model.

[0111] Step S503: The data in the training set and the test set is normalized.

[0112] Step S504: A full connection neural network is used to build a model architecture. The feature parameters input to the model are the clothing level, the body sensation wind speed, the background radiation temperature, the environment temperature, and the environment humidity, a total of n1. The input layer has n1 neurons. The first hidden layer has n2 neurons, and the second hidden layer has n3 neurons. The output layer has n4 neurons, which are exemplarily corresponding to 9 target labels in step S. The activation function used between the first n5 layers of the neural network can be ReLU, and the activation function used between the second hidden layer and the output layer is Sigmoid. The number of n1, n2, n3, n4, and n5 can be set according to actual use needs.

[0113] Step S505: The feature parameters of the training set are input to the neural network input layer once to train the model and obtain a prediction result. The prediction result and the target parameter are input into a cross entropy loss function to calculate the loss, and the model parameters are optimized through an Adam optimizer.

[0114] Step S506: The test set is input into the trained full connection neural network model to obtain a prediction result. The prediction result is compared with the target parameter to obtain the prediction accuracy of the model. By adjusting the model parameters, a verified prediction model is obtained.

[0115] Step S507: The verified prediction model is deployed. The clothing level, the body sensation wind speed, the background radiation temperature, the environment temperature, and the environment humidity of the human target detected in the use state of the prediction model input are obtained. n4 output values (i.e., classification probability values) are obtained. The integer index corresponding to the maximum probability value is the output human target comfort level category. Exemplarily, n4 can be 9.

[0116] The air conditioning equipment 10 further includes a control module configured to output a control parameter based on the user comfort level category and the air supply mode output by the prediction model. The control parameter includes at least one of a target temperature, a target humidity, a set wind speed, a set wind direction, a condensing temperature, an evaporating temperature, a throttling element opening degree, and a compressor speed.

[0117] Exemplarily, if the air supply mode is wind-avoiding-blowing. If the user comfort level category output by the prediction model is integer index 8, it means that the apparent temperature is appropriate, the apparent humidity is appropriate, and the user feels comfortable, so the current air conditioning operation mode is kept running. If the user comfort level category output by the prediction model is one of integer indexes 0-2, the target temperature is raised; if the user comfort level category output by the prediction model is one of integer indexes 3-5, the target temperature is lowered; if the user comfort level category output by the prediction model is one of integer indexes 0, 3, 6, the humidification module is turned on; if the user comfort level category output by the prediction model is one of 1, 4, 7, the dehumidification mode is run.

[0118] If the air supply mode is non-wind-avoiding-blowing. If the user comfort level category output by the prediction model is integer index 8, it means that the apparent temperature is appropriate, the apparent humidity is appropriate, and the user feels comfortable, so the current air conditioning operation mode is kept running. If the user comfort level category output by the prediction model is one of integer indexes 0-2, the target temperature is raised; if the user comfort level category output by the prediction model is one of integer indexes 3-5, the target temperature is lowered; if the user comfort level category output by the prediction model is one of integer indexes 0, 3, 6, the humidification module is turned on; if the user comfort level category output by the prediction model is one of 1, 4, 7, the dehumidification mode is run.

[0119] The prediction model runs once every set decision period, which is in units of minutes and can be set according to actual needs.

[0120] The air conditioning equipment 10 can be obtained by function expansion of the existing air conditioning equipment 10 in the installed and used state. In terms of hardware, an infrared temperature imaging sensor 111 is added to the air conditioning equipment 10 in use, which can form an indoor environment thermal image, so as to obtain the clothing level, the apparent wind speed, the indoor environment background radiation temperature, the environment temperature and the environment humidity of the user as the test human target. By collecting the user interaction of the user with the air conditioning room through the wire control, the remote control and the mobile control terminal, the user's apparent temperature comfort level and apparent humidity comfort level are obtained, and each combination of the apparent temperature comfort level and the apparent comfort level is assigned a corresponding integer index to represent the target label of each combination. Since the data of the air conditioning equipment 10 is closer to the use scene, the use habits and preferences of the user can be better reflected, so that a more accurate and reliable prediction model can be trained according to a small sample (several hundred) data.

[0121] However, the thermal comfort of the user in a certain room is not completely fixed. The thermal comfort is affected by multiple factors and can change over time, season, personnel, etc. In actual use scenarios, the small sample obtained can not be rich and diverse enough, so that the trained prediction model cannot be well generalized to other scenarios, resulting in control failure.

[0122] To solve this problem, in one or more embodiments of the present application, the control module is further configured to: obtain the integer index corresponding to the maximum probability value and the integer index corresponding to the second maximum probability value output by the prediction model, and calculate the difference between the maximum probability value and the second maximum probability value. When the difference is lower than a preset difference threshold, the output control parameter is suspended for at least one decision period, and the current running parameter is kept unchanged.

[0123] When the difference between the maximum probability value and the second maximum probability value is small, it indicates that the output result is near the decision boundary between the two indexes, i.e. a small disturbance or noise can cause the change of the classification result. In this embodiment, the large probability is caused by the insufficient training data of small sample data in the two categories, for example, the training data is collected in summer, and after the season is changed, the data imbalance affects the judgment of the prediction model. At this time, it is preferred to keep the current running parameter unchanged to help improve the accuracy and stability of the system.

[0124] The prediction difference value can be set according to the actual situation, for example, set to 0.2. For example, when the prediction model trained by small sample is configured, if the probability of integer index 8 is 0.45 (the maximum probability value), the probability of integer index 7 is 0.4 (the second maximum probability value), and the probability of integer index 6 is 0.15. The difference between the maximum probability value and the second maximum probability value is only 0.05. The prediction model is not clear for the distinction between integer index 8 and integer index 7, which is easy to cause false prediction. At this time, the air conditioning equipment 10 suspends the output of the control parameter for at least one decision period.

[0125] If the difference between the maximum probability value and the second maximum probability value is lower than the preset difference threshold for a plurality of times in succession, the control module triggers the first feature parameter acquisition module 102, the second feature parameter acquisition module 103, the target parameter acquisition module 104 to re-collect samples, and triggers the prediction module 105 to re-train. Thus, a feedback mechanism is provided, so that the prediction model better adapts to the current use environment and user habits. The newly collected samples can be combined with the previous samples to form a new sample set, improve the generalization ability and prediction accuracy of the prediction model, and dynamically adjust itself to cope with new situations and changes, improve flexibility and response ability.

[0126] In the description of the foregoing embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0127] The above merely describes specific implementations of the present application, but the protection scope of the present application is not limited thereto, any changes or replacements within the technical scope disclosed by the present application, which can be easily thought by any person skilled in the art, should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An air conditioning device, comprising: a sensing module comprising an infrared temperature imaging sensor configured to form a thermal image of an indoor environment; characterized in that it further comprises: a first characteristic parameter acquisition module configured to acquire, based on the thermal image, a clothing level of a test human target, a felt air speed, and a background radiation temperature of the indoor environment; a second characteristic parameter acquisition module configured to acquire an ambient temperature and an ambient humidity of the indoor environment; a target parameter acquisition module configured to acquire a combination of a felt temperature and a felt humidity of the test human target, and to represent a target label of each combination with an integer index; and a prediction module having a prediction model generated based on the clothing level, the felt air speed, the background radiation temperature, the ambient temperature, the ambient humidity, and the target label, the prediction model outputting an integer index corresponding to a maximum probability value as an output human target comfort level.

2. The air conditioning device according to claim 1, characterized in that: the first characteristic parameter acquisition module is configured to identify and select the test human target in the thermal image of the indoor environment, and to determine the felt air speed corresponding to the air supply mode according to the air supply mode; if the air supply mode is a wind-avoiding-human-blowing mode, the felt air speed is a preset minimum felt air speed; if the air supply mode is a non-wind-avoiding-human-blowing mode, then: identifying a relative distance and a relative angle between the test human target and an air outlet; estimating the felt air speed based on the relative distance, the relative angle, and a real-time air supply speed.

3. The air conditioning device according to claim 2, characterized in that: the first characteristic parameter acquisition module is configured to determine a target temperature of each pixel point in the human infrared thermal image, to screen a maximum pixel temperature value in the target temperature, and to calculate an average pixel temperature value of the human infrared thermal image; when a proportional relationship between the average pixel temperature value and the maximum pixel temperature value satisfies one of a plurality of preset clothing amount ratio threshold conditions, it is determined that a clothing amount of the test human target is a clothing level corresponding to the clothing amount ratio threshold condition.

4. The air conditioning device according to claim 3, characterized in that: the first characteristic parameter acquisition module is configured to filter out the human infrared thermal image in the thermal image of the indoor environment, and to generate the background radiation temperature based on the thermal image of the indoor environment after filtering out the human infrared thermal image.

5. The air conditioning device according to claim 4, characterized in that: the second characteristic parameter acquisition module classifies the ambient temperature based on preset temperature levels, and outputs an ambient temperature level.

6. The air conditioning device according to claim 5, characterized in that: the second characteristic parameter acquisition module classifies the ambient humidity based on preset humidity levels, and outputs an ambient humidity level.

7. The air conditioning device according to claim 6, characterized in that: the target parameter acquisition module is configured to: acquire a felt temperature comfort level corresponding to a felt temperature of the test human target; acquire a felt humidity comfort level corresponding to a felt humidity of the test human target; A corresponding integer index is assigned to each combination of the individual thermal comfort level and the individual humidity comfort level to represent the target label of each combination. 8.The air conditioning device of claim 7, wherein, The target parameter acquisition module is configured to: represent the thermal comfort level of the test human target by a thermal comfort level vector, wherein one bit of the thermal comfort level vector is 1 and the other bits are 0; represent the humidity comfort level of the test human target by a humidity comfort level vector, wherein one bit of the humidity comfort level vector is 1 and the other bits are 0; convert each combination of the thermal comfort level vector and the humidity comfort level vector into an integer index, wherein the value of the integer index represents the target label of each combination.

9. The air conditioning apparatus according to any one of claims 1 to 8, characterized by Further comprising: a control module configured to output a control parameter based on the user comfort category and the air supply mode output by the prediction model; the control parameter comprises at least one of a target temperature, a target humidity, a set air speed, a set air direction, a condensing temperature, an evaporating temperature, a throttling element opening degree, and a compressor rotating speed. 10.The air conditioning device of claim 9, wherein, the control module is further configured to acquire an integer index corresponding to the maximum probability value and an integer index corresponding to the second maximum probability value output by the prediction model, and calculate the difference between the maximum probability value and the second maximum probability value, and when the difference is lower than a preset difference threshold, suspend the output of the control parameter for at least one determination period and keep the current operating parameter unchanged.

Images