Air conditioner comfort model training method and device based on use habits of users

Abstract

The invention discloses a user use habit-based air conditioner comfort model training method and device, and the method comprises the steps: analyzing habit control parameters of a user in a target scene according to a determined first control data set of the user; the habit control parameters are input into a comfort model to be trained, a target comfort model trained to be converged is obtained, the comfort model is determined according to the target scene, and the target comfort model is used for generating target control parameters meeting the comfort requirement of the user in the target scene. Visibly, according to the embodiment of the invention, the training convergence accuracy of the target comfort model and the accuracy of outputting the target control parameters by the target comfort model can be improved by analyzing the use habits of the user and combining the target scene where the user is located, so that the user differentiation regulation control of the air conditioner is realized; and the air conditioner using comfort of a user can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and device for training an air-conditioning comfort model based on user usage habits. Background Art

[0002] With the development of science and technology and the improvement of people's living standards, air conditioning has gradually penetrated into every corner of people's lives. By sending temperature control instructions, people can obtain corresponding temperature compensation and humidity compensation at the place where the air conditioner is located to meet the user's needs.

[0003] In practical applications, in addition to the basic air conditioning temperature and humidity compensation performance, which is a factor that affects the user's choice of air conditioning, the comfort of air conditioning use is also an important factor that affects the user's choice of air conditioning. The prior art usually adopts the following method to improve the user's comfort of air conditioning use, that is, real-time detection of ambient temperature data, and comparison of the ambient temperature data with a preset threshold value to obtain a comparison result, and adjust the control scheme of the air conditioning according to the comparison result, wherein the control scheme is generally preset based on expert experience, however, this does not take into account the individual differences of users and actual environmental differences, and cannot perform adaptive adjustment of the air conditioning according to the actual state of each user, thereby failing to meet the user's comfort requirements for air conditioning use.

[0004] It can be seen that how to achieve user-differentiated adjustment and control of air conditioners to improve users' comfort in using air conditioners is particularly important. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an air conditioning comfort model training method and device based on user usage habits, which can be beneficial to realize user-differentiated adjustment and control of the air conditioner to improve the user's air conditioning comfort.

[0006] In order to solve the above technical problems, the first aspect of the present invention discloses a method for training an air conditioning comfort model based on user usage habits, the method comprising:

[0007] Analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user;

[0008] The habitual control parameters are input into a comfort model for training to obtain a target comfort model that is trained to convergence. The comfort model is determined based on the target scenario. The target comfort model is used to generate target control parameters that meet the comfort requirements of the user in the target scenario.

[0009] As an optional implementation manner, in the first aspect of the present invention, the first control data set includes a user identification data subset and a user instruction data subset, and the user instruction data subset is used to represent the comfort requirement of the user in the target scene;

[0010] Furthermore, before analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user, the method further includes:

[0011] According to the user identification data subset, determining whether the user is in the historical user set, when it is determined that the user is in the historical user set, determining a model matching the user instruction data subset in a pre-stored target comfort model set according to the user instruction data subset, and determining the model as the target comfort model;

[0012] When it is determined that the user is not in the historical user set, the operation of executing the first control data set based on the determined user, analyzing the user's habitual control parameters in the target scenario; inputting the habitual control parameters into the comfort model for training, and obtaining the target comfort model trained to convergence.

[0013] As an optional implementation manner, in the first aspect of the present invention, analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user includes:

[0014] Detecting a user data subset of the user and an environment data subset of a target scene where the user is located according to the determined first control data set of the user;

[0015] Analyzing the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset;

[0016] The user data subset includes at least one of the number and distribution data of individuals in the target scene, user gesture data, user voice data, user expression data, user face data, user body shape data, user metabolic rate data, user weight-to-height ratio, and user temperature sensitivity coefficient;

[0017] The environmental data subset includes at least one of spatial heat source data, object layout data, spatial pattern data, spatial wind field data, spatial temperature data, spatial humidity data, spatial noise field data, and spatial edge area data.

[0018] As an optional implementation, in the first aspect of the present invention, before determining, according to the user instruction data subset, a model matching the user instruction data subset in a pre-stored target comfort model set, the method further includes:

[0019] Determining whether there is a target data subset matching the user instruction data subset in the second control data set;

[0020] When it is determined that the target data subset exists in the second control data set, the operation of determining, according to the user instruction data subset, a model matching the user instruction data subset in the pre-stored target comfort model set is triggered;

[0021] And, the method further comprises:

[0022] When it is determined that the target data subset does not exist in the second control data set, the operation of analyzing the user's habitual control parameters in the target scenario based on the determined first control data set is triggered; the habitual control parameters are input into the comfort model for training to obtain the target comfort model that is trained to convergence.

[0023] As an optional embodiment, in the first aspect of the present invention, the method further comprises:

[0024] Inputting the first control data set and the second control data set into a preset value assessment model to obtain a difference value score of the first control data set;

[0025] Determining first transaction information according to the difference value score and a preset transaction conversion algorithm;

[0026] determining second transaction information according to the user identification data subset;

[0027] According to the first transaction information and the second transaction information, a corresponding cash back transaction operation is performed on the user.

[0028] As an optional implementation manner, in the first aspect of the present invention, analyzing the habitual control parameters of the user in the target scenario according to the user data subset and the environment data subset includes:

[0029] Determining current physical state information of the user according to the user data subset;

[0030] Determining the current environment status information of the user according to the environment data subset;

[0031] Analyzing appropriate control parameters of the user in the target scenario according to the current body state information and the current environment state information;

[0032] The suitable control parameters are adjusted according to the user instruction data subset to obtain the user's customary control parameters in the target scenario.

[0033] As an optional implementation, in the first aspect of the present invention, the step of inputting the habit control parameter into a comfort model for training to obtain a target comfort model trained to convergence comprises:

[0034] According to preset model training conditions, the habit control parameters are divided into a training sample data set and a verification sample data set, and the comfort model is divided into a first sub-model set and a second sub-model, wherein the first sub-model set includes at least two basic models;

[0035] Input each training sample data in the training sample data set into the corresponding basic model respectively, to obtain a first recommendation result for each training sample data and a first target model set corresponding to the first sub-model set;

[0036] Inputting the first recommendation results of all the training sample data into the second sub-model to obtain the second recommendation results corresponding to all the training sample data and the second target model corresponding to the second sub-model, wherein the second prediction result includes the first target control parameters corresponding to all the training sample data;

[0037] Determining, according to the second prediction results corresponding to all the training sample data, the first target model set and the second target model as comfort models to be verified;

[0038] Inputting all verification sample data in the verification sample data set into the prediction model to be verified, and obtaining verification results corresponding to all the verification sample data, wherein the verification results include second target control parameters corresponding to all the verification sample data;

[0039] It is determined whether the verification result satisfies a preset verification completion condition. When it is determined that the verification result satisfies the preset verification completion condition, it is determined that the comfort model to be verified is a target comfort model trained to convergence.

[0040] The second aspect of the present invention discloses an air conditioning comfort model training device based on user usage habits, the device comprising:

[0041] An analysis module, configured to analyze the user's habitual control parameters in a target scenario according to the determined first control data set of the user;

[0042] A training module is used to input the habitual control parameters into a comfort model for training to obtain a target comfort model that is trained to convergence. The comfort model is determined based on the target scenario. The target comfort model is used to generate target control parameters that meet the comfort requirements of the user in the target scenario.

[0043] As an optional implementation, in the second aspect of the present invention, the first control data set includes a user identification data subset and a user instruction data subset, and the user instruction data subset is used to represent the comfort requirement of the user in the target scene;

[0044] And, the device further comprises:

[0045] a judgment module, used for judging whether the user is in a historical user set according to the user identification data subset before the analysis module analyzes the user's habitual control parameters in a target scenario according to the determined first control data set of the user, and triggering the analysis module to perform the operation of analyzing the user's habitual control parameters in a target scenario according to the determined first control data set of the user when it is judged that the user is not in the historical user set; and the training module to perform the operation of inputting the habitual control parameters into a comfort model for training to obtain a target comfort model trained to convergence;

[0046] A determination module is used to determine a model that matches the user instruction data subset in a pre-stored target comfort model set based on the user instruction data subset when the judgment module determines that the user is in the historical user set, and determine the model as the target comfort model.

[0047] As an optional implementation, in the second aspect of the present invention, the specific manner in which the analysis module analyzes the user's habitual control parameters in the target scenario according to the determined first control data set of the user includes:

[0048] Detecting a user data subset of the user and an environment data subset of a target scene where the user is located according to the determined first control data set of the user;

[0049] Analyzing the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset;

[0050] The user data subset includes at least one of the number and distribution data of individuals in the target scene, user gesture data, user voice data, user expression data, user face data, user body shape data, user metabolic rate data, user weight-to-height ratio, and user temperature sensitivity coefficient;

[0051] The environmental data subset includes at least one of spatial heat source data, object layout data, spatial pattern data, spatial wind field data, spatial temperature data, spatial humidity data, spatial noise field data, and spatial edge area data.

[0052] As an optional embodiment, in the second aspect of the present invention, the judgment module is also used to judge whether there is a target data subset matching the user instruction data subset in the second control data set before the determination module determines the model matching the user instruction data subset in the pre-stored target comfort model set based on the user instruction data subset; when it is determined that the target data subset exists in the second control data set, the determination module is triggered to execute the operation of determining the model matching the user instruction data subset in the pre-stored target comfort model set based on the user instruction data subset; when it is determined that the target data subset does not exist in the second control data set, the analysis module is triggered to execute the operation of analyzing the user's habitual control parameters in the target scenario based on the determined first control data set of the user; and the training module executes the operation of inputting the habitual control parameters into the comfort model for training to obtain the target comfort model trained to convergence.

[0053] As an optional implementation, in the second aspect of the present invention, the device further includes:

[0054] A calculation module, used for inputting the first control data set and the second control data set into a preset value evaluation model to obtain a difference value score of the first control data set;

[0055] The determination module is further used to determine the first transaction information according to the difference value score and a preset transaction conversion algorithm;

[0056] The determination module is further configured to determine second transaction information according to the user identification data subset;

[0057] An execution module is used to perform a corresponding cash back transaction operation for the user according to the first transaction information and the second transaction information.

[0058] As an optional implementation, in the second aspect of the present invention, the specific manner in which the analysis module analyzes the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset includes:

[0059] Determining current physical state information of the user according to the user data subset;

[0060] Determining the current environment status information of the user according to the environment data subset;

[0061] Analyzing appropriate control parameters of the user in the target scenario according to the current body state information and the current environment state information;

[0062] The suitable control parameters are adjusted according to the user instruction data subset to obtain the user's customary control parameters in the target scenario.

[0063] As an optional implementation, in the second aspect of the present invention, the training module inputs the habit control parameter into the comfort model for training, and the specific method of obtaining the target comfort model trained to convergence includes:

[0064] According to preset model training conditions, the habit control parameters are divided into a training sample data set and a verification sample data set, and the comfort model is divided into a first sub-model set and a second sub-model, wherein the first sub-model set includes at least two basic models;

[0065] Input each training sample data in the training sample data set into the corresponding basic model respectively, to obtain a first recommendation result for each training sample data and a first target model set corresponding to the first sub-model set;

[0066] Inputting the first recommendation results of all the training sample data into the second sub-model to obtain the second recommendation results corresponding to all the training sample data and the second target model corresponding to the second sub-model, wherein the second prediction result includes the first target control parameters corresponding to all the training sample data;

[0067] Determining, according to the second prediction results corresponding to all the training sample data, the first target model set and the second target model as comfort models to be verified;

[0068] Inputting all verification sample data in the verification sample data set into the prediction model to be verified, and obtaining verification results corresponding to all the verification sample data, wherein the verification results include second target control parameters corresponding to all the verification sample data;

[0069] It is determined whether the verification result satisfies a preset verification completion condition. When it is determined that the verification result satisfies the preset verification completion condition, it is determined that the comfort model to be verified is a target comfort model trained to convergence.

[0070] The third aspect of the present invention discloses another air conditioning comfort model training device based on user usage habits, the device comprising:

[0071] A memory storing executable program code;

[0072] a processor coupled to the memory;

[0073] The processor calls the executable program code stored in the memory to execute the air conditioning comfort model training method based on user usage habits disclosed in the first aspect of the present invention.

[0074] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the air conditioning comfort model training method based on user usage habits disclosed in the first aspect of the present invention.

[0075] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0076] In an embodiment of the present invention, the user's habitual control parameters in the target scene are analyzed based on the determined first control data set of the user; the habitual control parameters are input into the comfort model for training to obtain a target comfort model trained to convergence, the comfort model is determined based on the target scene, and the target comfort model is used to generate target control parameters that meet the user's comfort requirements in the target scene. It can be seen that the implementation of the embodiment of the present invention can analyze the user's habitual control parameters in the target scene based on the user's first control data set, the habitual control parameters are used to represent the user's usage habits in the target scene, and the habitual control parameters are input into the comfort model determined based on the above-mentioned target scene for training to obtain a target comfort model, thereby being able to improve the accuracy of the training convergence of the target comfort model and the accuracy of the target control parameters output by the target comfort model by analyzing the user's usage habits and combining the user's target scene, so as to achieve user-differentiated adjustment and control of the air conditioner, thereby facilitating the improvement of the user's comfort in using the air conditioner. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0078] Figure 1 It is a scene schematic diagram of an air conditioning comfort model training scene based on user usage habits disclosed in an embodiment of the present invention;

[0079] Figure 2 It is a flow chart of an air conditioning comfort model training method based on user usage habits disclosed in an embodiment of the present invention;

[0080] Figure 3 It is a flow chart of another method for training an air conditioning comfort model based on user usage habits disclosed in an embodiment of the present invention;

[0081] Figure 4 It is a structural schematic diagram of an air conditioning comfort model training device based on user usage habits disclosed in an embodiment of the present invention;

[0082] Figure 5 It is a structural schematic diagram of another air conditioning comfort model training device based on user usage habits disclosed in an embodiment of the present invention;

[0083] Figure 6 It is a structural schematic diagram of another air conditioning comfort model training device based on user usage habits disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0085] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or ends.

[0086] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0087] The present invention discloses a method and device for training an air conditioning comfort model based on user usage habits, which can analyze the user's habitual control parameters in a target scene according to the user's first control data set, and the habitual control parameters are used to represent the user's usage habits in the target scene, and the habitual control parameters are input into a comfort model determined according to the above target scene for training to obtain a target comfort model, thereby improving the accuracy of the training convergence of the target comfort model and the accuracy of the target control parameters output by the target comfort model by analyzing the user's usage habits and combining the user's target scene, so as to realize user-differentiated adjustment and control of the air conditioner, thereby facilitating improving the user's comfort in using the air conditioner. The following are detailed descriptions.

[0088] In order to better understand the air conditioning comfort model training method and device based on user usage habits described in the present invention, firstly, a scenario applicable to the air conditioning comfort model training method based on user usage habits is described. Specifically, the scenario schematic diagram can be as follows: Figure 1 As shown, Figure 1 FIG. 1 is a schematic diagram of a scenario in an embodiment of the present invention. Figure 1 As shown, the scene includes a space body (such as a house) divided into three areas, a smart air conditioner, and a user / target, wherein the three areas represent three target scenes respectively, and the user / target can perform different activities in the three areas to present different target scenes, and the target scene may include at least one of a user reading scene, a user sleeping scene, a user cooking scene, a user sports scene, and a user washing scene. When the terminal device (such as an air conditioner) to which the present invention is applied receives the first control data set of the user / target, the user's target scene can be determined according to the sending direction of the first control data set, and the user's habitual control parameters in the target scene are analyzed, and the comfort model is trained based on the habitual control parameters to obtain a target comfort model trained to convergence, wherein the comfort model is determined according to the target scene, and the target comfort model is used to generate target control parameters that meet the user's comfort requirements in the target scene. The above-mentioned target scene can be obtained by analyzing the communication perception signal transmitted under the above-mentioned space body and received under the space body, and the above-mentioned communication perception signal includes a communication perception fusion frame structure. The synaesthesia fusion frame structure includes a communication control domain (C-Ctr) and a perception control domain (S-Str). The above-mentioned target scene can also be determined by the user by controlling the corresponding communication perception control device. Specifically, it can be determined by the user through the APP setting of the corresponding sensor perception.

[0089] The determined synaesthesia fusion frame structure proposed by the present invention is mainly a situation where the communication control domain and the perception control domain are evenly distributed; at the same time, the scenes involved in the present invention may include smart home target detection scenes, vehicle-mounted intelligent network target detection scenes, emergency rescue target detection scenes, prevention and control target detection scenes, attendance punching target detection scenes, etc. It should be noted that Figure 1 The scene diagram shown is only for showing a scene applicable to the air conditioning comfort model training method based on user usage habits. The space body (such as a house), smart air conditioner, user / target, etc. involved are only shown schematically. Figure 1 The scene schematic diagram shown does not limit this.

[0090] The following is a detailed description of an air conditioning comfort model training method and device based on user usage habits.

[0091] Embodiment 1

[0092] See also Figure 2 , Figure 2 is a flow chart of a method for training an air conditioning comfort model based on user usage habits disclosed in an embodiment of the present invention. Figure 2 The described air conditioning comfort model training method based on user usage habits can be applied to air conditioners, and can also be applied to terminal devices / cloud platforms that communicate / relay communications / control air conditioners and / or sensing devices, wherein the above-mentioned terminal devices include but are not limited to one or more of smart home devices, smart network devices, mobile communication devices, attendance punching devices, prevention and control equipment, and emergency rescue equipment, and the embodiments of the present invention do not limit this. Figure 2 As shown, the air conditioning comfort model training method based on user usage habits may include the following operations:

[0093] 101. Analyze the user's habitual control parameters in a target scenario according to the determined first control data set of the user.

[0094] In an embodiment of the present invention, the above-mentioned target scene is used to represent an application scenario for performing a target action in the target space; the above-mentioned target scene may include at least one of a user cooking scene, a user exercising scene, a user reading scene, a user watching a movie scene, a user washing scene, and a user sleeping scene. The above-mentioned target scene may be obtained by analyzing the communication perception signal transmitted in the target space and received in the target space, and the above-mentioned communication perception signal includes a communication perception fusion frame structure, or the above-mentioned target scene may be determined by the user by controlling the corresponding communication perception control device, specifically, it may be determined by the user through the corresponding sensor perception setting of the APP.

[0095] In the embodiment of the present invention, as an optional implementation, the first control data set includes a user identification data subset and a user instruction data subset, and the user instruction data subset is used to represent the user's comfort requirements in the target scene.

[0096] Optionally, the above-mentioned user identification data subset may include one or more of the user's facial data, voice data, fingerprint data, body data (including but not limited to one or more of gestures, postures, expressions), DNA data, personal profile data (including but not limited to one or more of name, age, height, blood type, skin color, skin type, hair quality, ID account, personal document account), or one or more label data assigned to the user based on the above-mentioned data.

[0097] In addition, the above-mentioned user command data subset may optionally include one or more of the user's voice data, fingerprint data, and body data (including but not limited to one or more of gestures, postures, and expressions); it may also be obtained by triggering the collection based on the above-mentioned user identification data subset. For example, the user sends the above-mentioned first control data set and sends it in the form of a signal. When the signal byte corresponding to the above-mentioned user identification data subset is determined, the signal byte corresponding to the above-mentioned user command data subset or the signal byte corresponding to the triggering collection of the user command data subset is read according to the preset signal byte specification to determine the user command data subset; it may also be one or more of the specific direct blowing control parameters, swing wind control parameters, set temperature parameters, etc. sent directly by the user.

[0098] The above-mentioned user identification data subset and user instruction data subset may also be set by the user through the above-mentioned APP.

[0099] Furthermore, before analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user, the method may further include the following operations:

[0100] Based on the user identification data subset, determine whether the user is in the historical user set. When it is determined that the user is in the historical user set, determine the model that matches the user instruction data subset in the pre-stored target comfort model set based on the user instruction data subset, and determine the model as the target comfort model.

[0101] When it is determined that the user is not in the historical user set, the execution is triggered to analyze the user's habitual control parameters in the target scenario based on the determined first control data set of the user; the habitual control parameters are input into the comfort model for training to obtain the operation of training to converge the target comfort model.

[0102] Among them, the above-mentioned target comfort model set can be stored in a server corresponding to the air conditioner or the relay communication device corresponding to the air conditioner. The above-mentioned target comfort model set is used to represent the target control parameters that are pre-trained and can generate to meet the comfort requirements of the corresponding user in the target scenario.

[0103] Optionally, the above-mentioned comfort model may include a recommendation-type algorithm model, such as a combination of one or more of linear regression, regression tree, deep learning, and nearest neighbor algorithm of the regression type, or a combination of one or more of K-means clustering, AP clustering, hierarchical clustering, and DBSCAN of the clustering type.

[0104] It can be seen that the implementation of this optional embodiment can determine whether the user is a historical user based on the user identification data subset of the user before analyzing the user's habitual control parameters in the target scenario based on the user's first control data set. On the one hand, if the user is a historical user, the model that matches the user instruction data subset of the user can be determined as the target comfort model in the model pool of the pre-stored target comfort model set, thereby improving the efficiency of determining the target comfort model, reducing unnecessary model training steps, saving hardware computing power, and preventing the model from overfitting or recommendation interference. On the other hand, if the user is a new user, the new needs of the new user can be used as a benchmark, and by analyzing the user's usage habits and combining the target scenario where the user is located, the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model can be improved, so as to achieve user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's air conditioning usage comfort.

[0105] In this optional embodiment, as an optional implementation manner, before determining the model matching the user instruction data subset in the pre-stored target comfort model set according to the user instruction data subset, the method may further include the following operations:

[0106] It is determined whether there is a target data subset matching the user instruction data subset in the second control data set.

[0107] When it is determined that the target data subset exists in the second control data set, an operation is triggered to determine, based on the user instruction data subset, an operation of determining a model matching the user instruction data subset in the pre-stored target comfort model set.

[0108] And, the method may also include the following operations:

[0109] When it is determined that the target data subset does not exist in the second control data set, the execution is triggered to analyze the user's habitual control parameters in the target scenario based on the determined first control data set of the user; the habitual control parameters are input into the comfort model for training to obtain the target comfort model that is trained to convergence.

[0110] Optionally, the second control data set is used to represent all historical users or a subset of user instruction data corresponding to the historical users.

[0111] It can be seen that the implementation of this optional embodiment can further determine whether there is a target data subset that matches the user instruction data subset of the historical user when it is determined that the user is a historical user. If there is a target data subset, it also indirectly indicates that there is a model that matches the target data subset and is in the target comfort model set, so that the model can be determined as the target comfort model. If there is no target data subset, the above-mentioned operations of analyzing user usage habits and training comfort models can continue to be triggered. This can further improve and ensure the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model on the basis of satisfying the efficiency of determining the target comfort model, so as to realize user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's comfort in using the air conditioner.

[0112] In this optional embodiment, as another optional implementation, the method may further include the following operations:

[0113] The first control data set and the second control data set are input into a preset value assessment model to obtain a difference value score of the first control data set.

[0114] The first transaction information is determined according to the difference value score and a preset transaction conversion algorithm.

[0115] The second transaction information is determined based on the user identification data subset.

[0116] According to the first transaction information and the second transaction information, a corresponding cash back transaction operation is performed for the user.

[0117] Optionally, the aforementioned user identification data subset may include user bank account information, or the user bank account information is indirectly obtained through other data in the aforementioned user identification data subset.

[0118] It can be seen that the implementation of this optional embodiment can, after determining the user's first control data set, combine the first control data set with the above-mentioned second control data set for processing to obtain the difference value score of the first control data set, so as to perform the corresponding cash back transaction operation to the user according to the preset transaction conversion algorithm and the user identification data subset, and provide the user with a certain model enhancement value contribution reward, which can further help to improve the user's participation and willingness in providing data, so as to improve the data richness and data quality of the target comfort model training, and further help to improve the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model.

[0119] In this optional embodiment, as an optional implementation, the preset value assessment model includes a first prediction model and a second prediction model, and the inputting of the first control data set and the second control data set into the preset value assessment model to obtain the difference value score of the first control data set may include the following operations:

[0120] The first control data set and the second control data set are input into the first prediction model to obtain the difference ratio of each difference between the first control data set and the second control data set.

[0121] For each difference, the difference ratio of the difference is input into the second prediction model to obtain an independent difference value score corresponding to the difference, and the independent difference value score is used to indicate the degree of improvement of the recommendation accuracy of the target comfort model by the difference.

[0122] Based on all of the independent difference value scores, a difference value score for the first control data set is determined.

[0123] It can be seen that the implementation of this optional embodiment can calculate the difference ratio of each difference existing in the first control data set relative to the second control data set, and further calculate the independent difference value score corresponding to each difference, which is beneficial to improving the calculation accuracy of the difference value score of the first control data set, and further helps to improve the calculation accuracy and fairness of providing rewards to users.

[0124] 102. The habit control parameters are input into the comfort model for training to obtain a target comfort model trained to convergence.

[0125] In the embodiment of the present invention, the above comfort model is determined according to the target scene, and the above target comfort model is used to generate target control parameters that meet the user's comfort requirements in the target scene.

[0126] It can be seen that the implementation of the embodiment of the present invention can analyze the user's habitual control parameters in the target scenario according to the user's first control data set, and the habitual control parameters are used to represent the user's usage habits in the target scenario, and the habitual control parameters are input into the comfort model determined according to the above-mentioned target scenario for training to obtain a target comfort model, thereby being able to improve the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model by analyzing the user's usage habits and combining the user's target scenario, so as to achieve user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's comfort in using the air conditioner.

[0127] Embodiment 2

[0128] See also Figure 3 , Figure 3 is a flow chart of a method for training an air conditioning comfort model based on user usage habits disclosed in an embodiment of the present invention. Figure 3 The described air conditioning comfort model training method based on user usage habits can be applied to air conditioners, and can also be applied to terminal devices / cloud platforms that communicate / relay communications / control air conditioners and / or sensing devices, wherein the above-mentioned terminal devices include but are not limited to one or more of smart home devices, smart network devices, mobile communication devices, attendance punching devices, prevention and control equipment, and emergency rescue equipment, and the embodiments of the present invention do not limit this. Figure 3 As shown, the air conditioning comfort model training method based on user usage habits may include the following operations:

[0129] 201. Detect a user data subset of the user and an environment data subset of a target scene where the user is located according to a determined first control data set of the user.

[0130] In an embodiment of the present invention, the above-mentioned first control data set can be used as a trigger instruction for triggering the above-mentioned operation of detecting the user data subset of the user and the environmental data subset of the target scene where the user is located, so that the device applied by the present invention knows how to perform the above-mentioned detection operation.

[0131] Specifically, in an embodiment of the present invention, as an optional implementation, the above-mentioned detecting the user data subset of the user and the environment data subset of the target scene where the user is located according to the determined first control data set of the user may include the following operations:

[0132] According to the determined first control data set of the user, the sending direction information of the first control data set is determined.

[0133] Based on the sending azimuth information, the target space and the spatial information of the target space are determined. The target space is used to represent the spatial area corresponding to the sending azimuth of the first control data set to be detected. The spatial information includes one or more of preset spatial pattern information, spatial area information, spatial light source information, and spatial function information.

[0134] Optionally, the above target space is included in a preset application space scenario.

[0135] A communication perception signal is determined according to the space information, and the communication perception signal is sent to the target space to detect a user data subset of the user and an environment data subset of the target scene where the user is located.

[0136] It can be seen that implementing this optional embodiment can indirectly determine the user's orientation information based on the user's first control data set, so as to further determine the target space to be detected, improve the efficiency and accuracy of perceiving the user data subset and the environmental data subset, and improve the accuracy of analyzing the user's habitual control parameters in the target scene.

[0137] Optionally, the user data subset and the environment data subset may be derived from each other;

[0138] For example, the user can set the household items corresponding to the user's frequent locations through a designated APP. When the layout position information of the household items is detected, the user's location information can be inferred preferentially. The above-mentioned household items may include but are not limited to beds, desks, chairs, sofas, etc.

[0139] Furthermore, when the user sets the identification information of the target space through a designated APP and associates it with a user data subset (personal information) of the user, then when the identification information of the target space is determined, the above-mentioned user data subset of the user can also be indirectly determined. For example, when it is detected that the identification information of the target space represents a user's bedroom, the above-mentioned user data subset stored in advance by the user can be retrieved.

[0140] 202. Analyze the user's habitual control parameters in the target scenario based on the user data subset and the environment data subset.

[0141] In the embodiment of the present invention, the relevant description of step 201 and step 202 can also refer to the specific description of step 101 in embodiment one, and the embodiment of the present invention will not go into details. On this basis, the embodiment of the present invention further supplements the description: in the embodiment of the present invention, the above-mentioned user data subset may include at least one of the number and distribution data of individuals in the target scene, user location data, user gesture data, user voice data, user expression data, user face data, user body shape data, user metabolic rate data, user weight-to-height ratio, and user temperature sensitivity coefficient.

[0142] The above-mentioned environmental data subset may include at least one of air conditioner location data, spatial heat source data, item layout data, spatial pattern data, spatial wind field data, spatial temperature data, spatial humidity data, spatial noise field data, and spatial edge area data. Optionally, the above-mentioned environmental data subset may also be set by the user through the above-mentioned APP.

[0143] It can be seen that the implementation of the embodiment of the present invention can perceive and detect the user's user data subset and the environmental data subset of the target scene where the user is located when determining the user's first control data set, and analyze the user's habitual control parameters in the target scene based on the perceived user data subset and environmental data subset, so as to improve the analysis accuracy of the user's habitual control parameters, which is beneficial to improving the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model, thereby helping to improve the user's air-conditioning comfort.

[0144] In the embodiment of the present invention, as an optional implementation, the above-mentioned analyzing the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset may include the following operations:

[0145] The user's current physical state information is determined based on the user data subset.

[0146] According to the subset of environmental data, the user's current environmental status information is determined.

[0147] According to the current body state information and the current environment state information, the appropriate control parameters for the user in the target scenario are analyzed.

[0148] According to the user instruction data subset, the appropriate control parameters are adjusted to obtain the user's habitual control parameters in the target scenario.

[0149] In this optional embodiment, the above-mentioned current physical state information includes the user's physical details comprehensively evaluated. Optionally, the physical details may refer to a physical examination report. Specifically, it may include but is not limited to at least one of human activity status information, human BMI index, human thermal map, and body surface liquid rate.

[0150] Similarly, the current environmental state information corresponds to the current body state information, for example, it may include but is not limited to at least one of the influence of the spatial edge area on the spatial temperature data, spatial humidity data, spatial wind field data, and spatial noise field data and the comparison of specific influence values.

[0151] It can be seen that the implementation of this optional embodiment can combine the user's current physical state information and the current environmental state information of the target scene where the user is located, and preferentially recommend a suitable control parameter that is suitable for the user in the target scene, and then re-adjust the suitable control parameter based on the user instruction data subset to obtain the user's habitual control parameter in the target scene. This can improve the accuracy of determining the habitual control parameter and save hardware computing power, which is conducive to training the comfort model based on the habitual control parameter to obtain the target comfort model.

[0152] In this optional embodiment, as an optional implementation manner, the above-mentioned analysis of the appropriate control parameters of the user in the target scene according to the current body state information and the current environment state information may include the following operations:

[0153] The current activity state of the user is analyzed according to the current body state information, and the current activity state is used to indicate the current activity type of the user, and the current activity type is used when the current activity is one or more of sleeping, reading, cooking, washing, and exercising.

[0154] Analyze the user's temperature sensitivity coefficient based on the user's current activity status.

[0155] According to the temperature sensitivity coefficient, appropriate control parameters for the user in the target scene that match the temperature sensitivity coefficient are determined.

[0156] It can be seen that the implementation of this optional embodiment can further determine the user's current activity state by analyzing the above-mentioned current body state information, and determine the user's temperature sensitivity coefficient based on the user's current activity state, so as to determine the user's appropriate control parameters in the target scenario. By determining the appropriate control parameters in combination with the user's actual activity state, the accuracy of determining the appropriate control parameters can be improved.

[0157] 203. The habit control parameters are input into the comfort model for training to obtain a target comfort model trained to convergence.

[0158] In the embodiment of the present invention, the above comfort model is determined according to the target scene, and the above target comfort model is used to generate target control parameters that meet the user's comfort requirements in the target scene.

[0159] In the embodiment of the present invention, as another optional implementation, the above-mentioned inputting the habit control parameter into the comfort model for training to obtain the target comfort model trained to convergence may include the following operations:

[0160] According to preset model training conditions, the habit control parameters are divided into a training sample data set and a verification sample data set, and the comfort model is divided into a first sub-model set and a second sub-model, wherein the first sub-model set includes at least two basic models.

[0161] Each training sample data in the training sample data set is input into the corresponding basic model respectively to obtain the first recommendation result of each training sample data and the first target model set corresponding to the first sub-model set.

[0162] The first recommendation results of all training sample data are input into the second sub-model to obtain the second recommendation results corresponding to all training sample data and the second target model corresponding to the second sub-model. The second prediction results include the first target control parameters corresponding to all training sample data.

[0163] According to the second prediction results corresponding to all the training sample data, the first target model set and the second target model are determined as the comfort models to be verified.

[0164] All verification sample data in the verification sample data set are input into the prediction model to be verified to obtain verification results corresponding to all verification sample data, and the verification results include the second target control parameters corresponding to all verification sample data.

[0165] It is determined whether the verification result satisfies a preset verification completion condition. When it is determined that the verification result satisfies the preset verification completion condition, it is determined that the comfort model to be verified is a target comfort model trained to convergence.

[0166] It can be seen that implementing this optional embodiment can provide a specific model architecture of the target comfort model and an implementation method for training and verifying the target comfort model. By combining the first target model set and the second target model to jointly recommend and generate target control parameters, it can be beneficial to improve the accuracy of recommendation generation of the target comfort model, thereby helping to improve the user's air-conditioning comfort.

[0167] Embodiment 3

[0168] See also Figure 4 , Figure 4 : is a schematic diagram of the structure of an air conditioning comfort model training device based on user usage habits disclosed in an embodiment of the present invention. Figure 4The described air conditioning comfort model training device based on user usage habits can be applied to air conditioners, and can also be applied to terminal devices / cloud platforms that communicate / relay communications / control with air conditioners and / or sensing devices, wherein the above-mentioned terminal devices include but are not limited to one or more of smart home devices, smart network devices, mobile communication devices, attendance punching devices, prevention and control equipment, and emergency rescue equipment, and the embodiments of the present invention do not limit this. Figure 4 As shown, the air conditioning comfort model training device based on user usage habits may include:

[0169] The analysis module 301 is used to analyze the user's habitual control parameters in the target scenario according to the determined first control data set of the user.

[0170] The training module 302 is used to input the habit control parameters into the comfort model for training, so as to obtain a target comfort model trained to convergence.

[0171] In the embodiment of the present invention, the above comfort model is determined according to the target scene, and the above target comfort model is used to generate target control parameters that meet the user's comfort requirements in the target scene.

[0172] It can be seen that the implementation of the embodiment of the present invention can analyze the user's habitual control parameters in the target scenario according to the user's first control data set, and the habitual control parameters are used to represent the user's usage habits in the target scenario, and the habitual control parameters are input into the comfort model determined according to the above-mentioned target scenario for training to obtain a target comfort model, thereby being able to improve the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model by analyzing the user's usage habits and combining the user's target scenario, so as to achieve user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's comfort in using the air conditioner.

[0173] In an embodiment of the present invention, as an optional implementation, the first control data set includes a user identification data subset and a user instruction data subset, and the user instruction data subset is used to represent the user's comfort requirements in the target scene.

[0174] And, if Figure 5 As shown, the device may also include:

[0175] The judgment module 303 is used to judge whether the user is in the historical user set according to the user identification data subset before the analysis module 301 analyzes the user's habitual control parameters in the target scenario according to the determined first control data set of the user. When it is judged that the user is not in the historical user set, the analysis module 301 is triggered to perform the operation of analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user; and the training module 302 performs the operation of inputting the habitual control parameters into the comfort model for training to obtain the target comfort model that is trained to convergence.

[0176] The determination module 304 is used to determine the model that matches the user instruction data subset in the pre-stored target comfort model set based on the user instruction data subset when the judgment module 303 determines that the user is in the historical user set, and determine the model as the target comfort model.

[0177] It can be seen that the implementation of this optional embodiment can determine whether the user is a historical user based on the user identification data subset of the user before analyzing the user's habitual control parameters in the target scenario based on the user's first control data set. On the one hand, if the user is a historical user, the model that matches the user instruction data subset of the user can be determined as the target comfort model in the model pool of the pre-stored target comfort model set, thereby improving the efficiency of determining the target comfort model, reducing unnecessary model training steps, saving hardware computing power, and preventing the model from overfitting or recommendation interference. On the other hand, if the user is a new user, the new needs of the new user can be used as a benchmark, and by analyzing the user's usage habits and combining the target scenario where the user is located, the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model can be improved, so as to achieve user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's air conditioning usage comfort.

[0178] In this optional embodiment, as an optional implementation, the specific manner in which the analysis module 301 analyzes the user's habitual control parameters in the target scenario according to the determined first control data set of the user includes:

[0179] According to the determined first control data set of the user, a user data subset of the user and an environment data subset of the target scene where the user is located are detected.

[0180] Based on the user data subset and the environment data subset, the user's habit control parameters in the target scenario are analyzed.

[0181] Among them, the user data subset includes at least one of the number and distribution data of individuals in the target scene, user gesture data, user voice data, user expression data, user face data, user body shape data, user metabolic rate data, user weight-to-height ratio, and user temperature sensitivity coefficient.

[0182] The environmental data subset includes at least one of spatial heat source data, object layout data, spatial pattern data, spatial wind field data, spatial temperature data, spatial humidity data, spatial noise field data, and spatial edge area data.

[0183] It can be seen that the implementation of the embodiment of the present invention can perceive and detect the user's user data subset and the environmental data subset of the target scene where the user is located when determining the user's first control data set, and analyze the user's habitual control parameters in the target scene based on the perceived user data subset and environmental data subset, so as to improve the analysis accuracy of the user's habitual control parameters, which is beneficial to improving the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model, thereby helping to improve the user's air-conditioning comfort.

[0184] In this optional embodiment, as another optional implementation, the above-mentioned judgment module 303 is also used to judge whether there is a target data subset matching the user instruction data subset in the second control data set before the determination module 304 determines the model matching the user instruction data subset in the pre-stored target comfort model set according to the user instruction data subset; when it is determined that the target data subset exists in the second control data set, the determination module 304 is triggered to execute the operation of determining the model matching the user instruction data subset in the pre-stored target comfort model set according to the user instruction data subset; when it is determined that the target data subset does not exist in the second control data set, the analysis module 301 is triggered to execute the operation of analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user; and the training module 302 executes the operation of inputting the habitual control parameters into the comfort model for training to obtain the target comfort model trained to convergence.

[0185] It can be seen that the implementation of this optional embodiment can further determine whether there is a target data subset that matches the user instruction data subset of the historical user when it is determined that the user is a historical user. If there is a target data subset, it also indirectly indicates that there is a model that matches the target data subset and is in the target comfort model set, so that the model can be determined as the target comfort model. If there is no target data subset, the above-mentioned operations of analyzing user usage habits and training comfort models can continue to be triggered. This can further improve and ensure the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model on the basis of satisfying the efficiency of determining the target comfort model, so as to realize user-differentiated adjustment and control of the air conditioner, which is beneficial to improving the user's comfort in using the air conditioner.

[0186] In the embodiment of the present invention, as another optional implementation, Figure 5 As shown, the device may also include:

[0187] The calculation module 305 is used to input the first control data set and the second control data set into a preset value assessment model to obtain a difference value score of the first control data set.

[0188] The determination module 304 is further configured to determine the first transaction information according to the difference value score and a preset transaction conversion algorithm.

[0189] The determination module 304 is further configured to determine the second transaction information according to the user identification data subset.

[0190] The execution module 306 is used to perform a corresponding cash back transaction operation for the user according to the first transaction information and the second transaction information.

[0191] It can be seen that the implementation of this optional embodiment can, after determining the user's first control data set, combine the first control data set with the above-mentioned second control data set for processing to obtain the difference value score of the first control data set, so as to perform the corresponding cash back transaction operation to the user according to the preset transaction conversion algorithm and the user identification data subset, and provide the user with a certain model enhancement value contribution reward, which can further help to improve the user's participation and willingness in providing data, so as to improve the data richness and data quality of the target comfort model training, and further help to improve the training convergence accuracy of the target comfort model and the accuracy of the target control parameters output by the target comfort model.

[0192] In an optional embodiment, the specific manner in which the analysis module 301 analyzes the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset includes:

[0193] The user's current physical state information is determined based on the user data subset.

[0194] According to the subset of environmental data, the user's current environmental status information is determined.

[0195] According to the current body state information and the current environment state information, the appropriate control parameters for the user in the target scenario are analyzed.

[0196] According to the user instruction data subset, the appropriate control parameters are adjusted to obtain the user's habitual control parameters in the target scenario.

[0197] It can be seen that the implementation of this optional embodiment can combine the user's current physical state information and the current environmental state information of the target scene where the user is located, and preferentially recommend a suitable control parameter that is suitable for the user in the target scene, and then re-adjust the suitable control parameter based on the user instruction data subset to obtain the user's habitual control parameter in the target scene. This can improve the accuracy of determining the habitual control parameter and save hardware computing power, which is conducive to training the comfort model based on the habitual control parameter to obtain the target comfort model.

[0198] In another optional embodiment, the training module 302 inputs the habit control parameters into the comfort model for training, and the specific method of obtaining the target comfort model trained to convergence includes:

[0199] According to preset model training conditions, the habit control parameters are divided into a training sample data set and a verification sample data set, and the comfort model is divided into a first sub-model set and a second sub-model, wherein the first sub-model set includes at least two basic models.

[0200] Each training sample data in the training sample data set is input into the corresponding basic model respectively to obtain the first recommendation result of each training sample data and the first target model set corresponding to the first sub-model set.

[0201] The first recommendation results of all training sample data are input into the second sub-model to obtain the second recommendation results corresponding to all training sample data and the second target model corresponding to the second sub-model. The second prediction results include the first target control parameters corresponding to all training sample data.

[0202] According to the second prediction results corresponding to all the training sample data, the first target model set and the second target model are determined as the comfort models to be verified.

[0203] All verification sample data in the verification sample data set are input into the prediction model to be verified to obtain verification results corresponding to all verification sample data, and the verification results include the second target control parameters corresponding to all verification sample data.

[0204] It is determined whether the verification result satisfies a preset verification completion condition. When it is determined that the verification result satisfies the preset verification completion condition, it is determined that the comfort model to be verified is a target comfort model trained to convergence.

[0205] It can be seen that implementing this optional embodiment can provide a specific model architecture of the target comfort model and an implementation method for training and verifying the target comfort model. By combining the first target model set and the second target model to jointly recommend and generate target control parameters, it can be beneficial to improve the accuracy of recommendation generation of the target comfort model, thereby helping to improve the user's air-conditioning comfort.

[0206] Embodiment 4

[0207] See also Figure 6 , Figure 6 is a structural schematic diagram of another air conditioning comfort model training device based on user usage habits disclosed in an embodiment of the present invention. Figure 6 As shown, the air conditioning comfort model training device based on user usage habits may include:

[0208] The memory 401 stores executable program codes.

[0209] A processor 402 is coupled to the memory 401 .

[0210] The processor 402 calls the executable program code stored in the memory 401 to execute the steps in the air conditioning comfort model training method based on user usage habits described in the first embodiment of the present invention or the second embodiment of the present invention.

[0211] Embodiment 5

[0212] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the air conditioning comfort model training method based on user usage habits described in Embodiment 1 or Embodiment 2 of the present invention.

[0213] Embodiment 6

[0214] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the air conditioning comfort model training method based on user usage habits described in Example 1 or Example 2.

[0215] The device embodiments described above are only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, i.e., they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.

[0216] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0217] Finally, it should be noted that the air conditioning comfort model training method and device based on user usage habits disclosed in the embodiment of the present invention only discloses the preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, ordinary technicians in this field should understand that it is still possible to modify the technical solutions recorded in the aforementioned embodiments, or to replace some of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for training an air conditioning comfort model based on user usage habits, characterized in that: The method comprises: Analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user; The habitual control parameters are input into a comfort model for training to obtain a target comfort model that is trained to convergence. The comfort model is determined based on the target scenario. The target comfort model is used to generate target control parameters that meet the comfort requirements of the user in the target scenario.

2. The air conditioning comfort model training method based on user usage habits according to claim 1 is characterized in that: The first control data set includes a user identification data subset and a user instruction data subset, and the user instruction data subset is used to represent the comfort requirement of the user in the target scene; Furthermore, before analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user, the method further includes: According to the user identification data subset, determining whether the user is in the historical user set, when it is determined that the user is in the historical user set, determining a model matching the user instruction data subset in a pre-stored target comfort model set according to the user instruction data subset, and determining the model as the target comfort model; When it is determined that the user is not in the historical user set, the operation of executing the first control data set based on the determined user, analyzing the user's habitual control parameters in the target scenario; inputting the habitual control parameters into the comfort model for training, and obtaining the target comfort model trained to convergence.

3. The air conditioning comfort model training method based on user usage habits according to claim 2 is characterized in that: The step of analyzing the user's habitual control parameters in the target scenario according to the determined first control data set of the user includes: Detecting a user data subset of the user and an environment data subset of a target scene where the user is located according to the determined first control data set of the user; Analyzing the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset; The user data subset includes at least one of the number and distribution data of individuals in the target scene, user gesture data, user voice data, user expression data, user face data, user body shape data, user metabolic rate data, user weight-to-height ratio, and user temperature sensitivity coefficient; The environmental data subset includes at least one of spatial heat source data, object layout data, spatial pattern data, spatial wind field data, spatial temperature data, spatial humidity data, spatial noise field data, and spatial edge area data.

4. The air conditioning comfort model training method based on user usage habits according to claim 2 is characterized in that: Before determining, based on the user instruction data subset, a model matching the user instruction data subset in a pre-stored target comfort model set, the method further includes: Determining whether there is a target data subset matching the user instruction data subset in the second control data set; When it is determined that the target data subset exists in the second control data set, the operation of determining, according to the user instruction data subset, a model matching the user instruction data subset in the pre-stored target comfort model set is triggered; And, the method further comprises: When it is determined that the target data subset does not exist in the second control data set, the operation of analyzing the user's habitual control parameters in the target scenario based on the determined first control data set is triggered; the habitual control parameters are input into the comfort model for training to obtain the target comfort model that is trained to convergence.

5. The air conditioning comfort model training method based on user usage habits according to claim 4 is characterized in that: The method further comprises: Inputting the first control data set and the second control data set into a preset value assessment model to obtain a difference value score of the first control data set; Determining first transaction information according to the difference value score and a preset transaction conversion algorithm; determining second transaction information according to the user identification data subset; According to the first transaction information and the second transaction information, a corresponding cash back transaction operation is performed on the user.

6. The air conditioning comfort model training method based on user usage habits according to claim 3 is characterized in that: The analyzing the user's habitual control parameters in the target scenario according to the user data subset and the environment data subset includes: Determining current physical state information of the user according to the user data subset; Determining the current environment status information of the user according to the environment data subset; Analyzing appropriate control parameters of the user in the target scenario according to the current body state information and the current environment state information; The suitable control parameters are adjusted according to the user instruction data subset to obtain the user's customary control parameters in the target scenario.

7. The air conditioning comfort model training method based on user usage habits according to any one of claims 1 to 6, characterized in that: The step of inputting the habit control parameter into the comfort model for training to obtain a target comfort model trained to convergence includes: According to preset model training conditions, the habit control parameters are divided into a training sample data set and a verification sample data set, and the comfort model is divided into a first sub-model set and a second sub-model, wherein the first sub-model set includes at least two basic models; Input each training sample data in the training sample data set into the corresponding basic model respectively, to obtain a first recommendation result for each training sample data and a first target model set corresponding to the first sub-model set; Inputting the first recommendation results of all the training sample data into the second sub-model to obtain the second recommendation results corresponding to all the training sample data and the second target model corresponding to the second sub-model, wherein the second prediction result includes the first target control parameters corresponding to all the training sample data; Determining, according to the second prediction results corresponding to all the training sample data, the first target model set and the second target model as comfort models to be verified; Inputting all verification sample data in the verification sample data set into the prediction model to be verified, and obtaining verification results corresponding to all the verification sample data, wherein the verification results include second target control parameters corresponding to all the verification sample data; It is determined whether the verification result satisfies a preset verification completion condition. When it is determined that the verification result satisfies the preset verification completion condition, it is determined that the comfort model to be verified is a target comfort model trained to convergence.

8. An air conditioning comfort model training device based on user usage habits, characterized in that: The device comprises: An analysis module, configured to analyze the user's habitual control parameters in a target scenario according to the determined first control data set of the user; A training module is used to input the habitual control parameters into a comfort model for training to obtain a target comfort model that is trained to convergence. The comfort model is determined based on the target scenario. The target comfort model is used to generate target control parameters that meet the comfort requirements of the user in the target scenario.

9. An air conditioning comfort model training device based on user usage habits, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the air conditioning comfort model training method based on user usage habits as described in any one of claims 1-7.

10. A computer storage medium, characterized in that: The computer storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the air conditioning comfort model training method based on user usage habits as described in any one of claims 1-7.

Images