Smart home system scene control methods, control devices, and storage media

Abstract

This invention discloses a scene control method, a scene control device, and a storage medium for a smart home system. The method includes: acquiring scene feature information of a target indoor space; sequentially determining the matching results between various preset scenes and the scene feature information in at least two preset scenes, obtaining at least two recognition results; determining a target scene among the at least two preset scenes based on the at least two recognition results; and controlling the operation of at least one smart appliance within the target indoor space according to the target scene. This invention aims to improve the accuracy of scene recognition and device control.

Description

Technical Field

[0001] This invention relates to the field of smart home technology, and in particular to a scene control method, a scene control device, and a storage medium for a smart home system. Background Technology

[0002] In many smart home applications, scene recognition is used to control device operation based on the identified scenes. Typically, smart home systems have multiple pre-set scenes. The system checks whether the monitoring parameters of the indoor environment meet the conditions for each scene in a certain order. Once a matching scene is identified, it stops identifying other scenes. This can easily lead to inaccurate scene recognition results, affecting the accuracy of device control. Summary of the Invention

[0003] The main objective of this invention is to provide a scene control method, a scene control device, and a storage medium for a smart home system, aiming to improve the accuracy of scene recognition and device control.

[0004] To achieve the above objectives, the present invention provides a scene control method for a smart home system, the scene control method for the smart home system comprising the following steps:

[0005] Obtain scene feature information of the target indoor space;

[0006] Sequentially determine the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results;

[0007] The target scene in the at least two preset scenarios is determined based on the at least two recognition results;

[0008] Control the operation of at least one smart appliance in the target indoor space according to the target scenario.

[0009] Optionally, the step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes:

[0010] Sequentially identify whether the preset scene information of each of the at least two preset scenarios matches the scene feature information, and obtain the at least two identification results.

[0011] Optionally, the step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes:

[0012] When there are user setting parameters corresponding to the preset scenario, the preset scenario information corresponding to the preset scenario is determined according to the user setting parameters, and the preset scenario information of each preset scenario in the at least two preset scenarios is sequentially identified to see if it matches the scenario feature information, so as to obtain the at least two identification results;

[0013] When no user setting parameters corresponding to the preset scenario exist, the scenario feature information is input into the scenario recognition model corresponding to each preset scenario to obtain the output result of each scenario recognition model. The recognition result includes the output result. The scenario recognition model is generated based on machine learning.

[0014] Optionally, the step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes:

[0015] The scene feature information is input into the scene recognition model corresponding to each preset scene, and the output result of each scene recognition model is obtained. The recognition result includes the output result. The scene recognition model is generated based on machine learning.

[0016] Optionally, the scene recognition model includes a first model or a second model, and the step of inputting the scene feature information into the scene recognition model corresponding to each preset scene and obtaining the output result corresponding to each scene recognition model includes:

[0017] The scene feature information is input into the first model corresponding to each preset scene, and a first output result is obtained for each first model. The first output result includes whether the current scene represented by the scene feature information is the corresponding preset scene.

[0018] Alternatively, the scene feature information can be input into the second model corresponding to each preset scene to obtain a second output result corresponding to each second model. The second output result includes the probability that the current scene represented by the scene feature information is the corresponding preset scene.

[0019] Optionally, the step of inputting the scene feature information into the scene recognition model corresponding to each preset scene and obtaining the output result corresponding to each scene recognition model includes:

[0020] Obtain the location type of the target indoor space;

[0021] When the location type is the first type, the scene feature information is input into the first model corresponding to each preset scene to obtain the first output result corresponding to each first model;

[0022] When the location type is the second type, the scene feature information is input into the second model corresponding to each preset scene to obtain the second output result corresponding to each second model;

[0023] Wherein, the first similarity between the at least two preset scenarios corresponding to the first type is less than the second similarity between the at least two preset scenarios corresponding to the second type.

[0024] Optionally, the step of determining the target scene among the at least two preset scenes based on the at least two recognition results includes:

[0025] When there is more than one recognition result that satisfies the preset conditions among the at least two recognition results, the target scene is determined according to the priority or predicted probability of all the preset scenes corresponding to all recognition results that satisfy the preset conditions.

[0026] Wherein, the preset condition indicates that the preset scene matches the scene feature information, and the prediction probability is the probability that the current scene represented by the scene feature information is the corresponding preset scene.

[0027] Optionally, the recognition result includes whether the current scene represented by the scene feature information is the corresponding preset scene, and the preset condition includes the recognition result being that the current scene represented by the scene feature information is the corresponding preset scene;

[0028] Alternatively, the identification result may include the predicted probability, and the preset condition may include the predicted probability being greater than the first preset probability.

[0029] Optionally, after the step of determining the target scene among the at least two preset scenes based on the at least two recognition results, the method further includes:

[0030] Output the switching prompt information corresponding to the target scene;

[0031] When a confirmation message corresponding to the target scenario is received, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scenario is executed.

[0032] Optionally, after the step of determining the target scene among the at least two preset scenes based on the at least two recognition results, the method further includes:

[0033] When the predicted probability corresponding to the target scenario is greater than the second preset probability, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scenario is executed.

[0034] When the predicted probability corresponding to the target scene is less than or equal to the second preset probability, the switching prompt information corresponding to the target scene is output. When the confirmation information corresponding to the target scene is received, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene is executed.

[0035] The predicted probability is the probability that the current scene, as represented by the scene feature information, is the target scene.

[0036] In addition, to achieve the above objectives, this application also proposes a scene control device for a smart home system, the scene control device for a smart home system comprising: a memory, a processor, and a scene control program for a smart home system stored in the memory and executable on the processor, wherein when the scene control program for a smart home system is executed by the processor, it implements the steps of the scene control method for a smart home system as described in any of the preceding claims.

[0037] In addition, to achieve the above objectives, this application also proposes a storage medium storing a scene control program for a smart home system, wherein when the scene control program for the smart home system is executed by a processor, it implements the steps of the scene control method for the smart home system as described in any of the preceding claims.

[0038] This invention proposes a scene control method for a smart home system. The method sequentially determines the matching status of scene feature information between each of at least two preset scenes and a target indoor space, obtaining at least two recognition results. Based on these two recognition results, a target scene is determined, and at least one smart appliance in the target indoor space is controlled according to the target scene. During this process, the recognition of the target scene comprehensively considers the matching status of all preset scenes, ensuring that no scene is unrecognized. This effectively improves the accuracy of scene recognition results and guarantees accurate control of smart appliances based on the recognized target scene, thereby effectively improving the accuracy of device control. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the hardware structure involved in the operation of an embodiment of the smart appliance of the present invention;

[0040] Figure 2 This is a flowchart illustrating an embodiment of the scene control method for the smart home system of the present invention;

[0041] Figure 3 This is a flowchart illustrating another embodiment of the scene control method for the smart home system of the present invention;

[0042] Figure 4 This is a flowchart illustrating another embodiment of the scene control method for the smart home system of the present invention;

[0043] Figure 5 This is a flowchart illustrating another optional embodiment of the scene control method for the smart home system of the present invention.

[0044] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0045] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0046] This invention provides a scene control device 1 for a smart home system.

[0047] In this embodiment of the invention, the scene control device 1 of the smart home system is applied to the smart home system, which includes one or more smart appliances 2 within the target space. (Refer to...) Figure 1 The scene control device 1 of the smart home system is connected to each smart appliance 2 in the smart home system.

[0048] In this embodiment, the target space is the entire house. In other embodiments, the target space may also be a combination of preset designated areas or a specific area, such as a room or a living room.

[0049] Among them, more than one smart appliance 2 can be connected to form a smart home system based on the connection module, which may include WIFI, Bluetooth, infrared, wiring, etc.

[0050] The smart home system may include a sensing module 3, which is connected to the scene control device 1 of the smart home system. The sensing module 3 can be used to receive sensing information. In this embodiment, the sensing module 3 may include environmental sensors, human body sensors, and device monitoring modules, etc., to detect sensing information such as environmental information, human body information, and device information.

[0051] Environmental sensors include sensors used to detect at least one environmental parameter such as temperature, humidity, PM2.5, carbon dioxide, and formaldehyde.

[0052] Human body sensors may include at least one type of sensor used to detect human information, such as infrared sensors, radar sensors, or WIFI. Human body information may include at least one type of information such as the number of people, their location, breathing, heart rate, and body movement amplitude.

[0053] The environmental information detected by the equipment monitoring module includes at least one of the following: power on / off status, set mode, set temperature, etc.

[0054] In this embodiment of the invention, the scene control device 1 of the smart home system can be a server set up independently of the smart appliance 2 and connected to the smart appliance 2, or it can be a functional module built into the smart appliance 2. (Refer to...) Figure 1 The scene control device 1 of the smart home system includes: a processor 1001, such as a CPU, a memory 1002, and a timer 1003. These components communicate with each other via a communication bus. The memory 1002 can be high-speed RAM or stable non-volatile memory, such as disk storage. Optionally, the memory 1002 can also be a storage device independent of the aforementioned processor 1001.

[0055] Those skilled in the art will understand that Figure 1 The device structure shown does not constitute a limitation on the device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0056] like Figure 1 As shown, the memory 1002, which serves as a computer storage medium, may include a scene control program for a smart home system.

[0057] exist Figure 1 In the device shown, the processor 1001 can be used to call the scene control program of the smart home system stored in the memory 1002 and execute the relevant steps of the scene control method of the smart home system in the following embodiments.

[0058] This invention also provides a scene control method for a smart home system, which is applied to the scene control device of the aforementioned smart home system.

[0059] Reference Figure 2 This application proposes an embodiment of a scene control method for a smart home system. In this embodiment, the scene control method for the smart home system includes:

[0060] Step S10: Obtain scene feature information of the target indoor space;

[0061] The target indoor space is an indoor space equipped with a smart home system.

[0062] Scene feature information refers to information related to the scene requirements within the target indoor space. This scene feature information can be detected by sensors within the target indoor space and / or collected by modules outside the target indoor space.

[0063] Scene feature information includes, but is not limited to, at least one of the following: environmental information, human body information, weather information, geographical information, and time information. Environmental information includes indoor environmental information (e.g., indoor temperature, indoor humidity, indoor PM2.5, etc.) and / or outdoor environmental information (e.g., outdoor temperature, outdoor humidity, outdoor PM2.5, etc.). Human body information includes at least one of the following: number of people, location of people, breathing, heart rate, and amplitude of body movement. Weather information includes sunny, rainy, foggy, or snowy weather. Geographical information includes country / region, province / city, geographical division (e.g., North, South, North China, Yellow River Basin, etc.) and climate zone (e.g., hot summer, warm winter, cold). Time information includes year / month, day / night time (e.g., morning or afternoon, day or night, etc.), and solar terms (e.g., White Dew, Minor Snow, etc.).

[0064] Step S20: Sequentially determine the matching status of various preset scenarios with the scene feature information in at least two preset scenarios to obtain at least two recognition results;

[0065] Preset scenarios can be user-defined or pre-configured by the user before the system leaves the factory. At least two preset scenarios may include, but are not limited to, at least two of the following: humid weather scenario, rainy season scenario, sleep scenario, dinner scenario, meeting scenario, etc.

[0066] Among them, at least two preset scenarios can be fixed scenarios that have been set in advance; or, at least two preset scenarios can be obtained according to the type of the target indoor space, with different types of the target indoor space corresponding to different at least two preset scenarios; or, at least two preset scenarios can be obtained from the types of all smart appliances in the target indoor space, and so on.

[0067] Matching conditions include the degree of matching or whether a match exists. Recognition results may include, but are not limited to, one of the following: the current scene represented by scene feature information matches a preset scene; the current scene represented by scene feature information does not match a preset scene; the current scene represented by scene feature information is a preset scene; the current scene represented by scene feature information is not a preset scene; the probability that the current scene represented by scene feature information is a preset scene, etc.

[0068] Preset scenarios can be configured with corresponding recognition rules (such as conditional recognition or algorithmic model recognition). In this embodiment, the recognition rules are the same for different preset scenarios. In other embodiments, the recognition rules for different preset scenarios may be different.

[0069] The scene feature information is analyzed according to the recognition rules corresponding to each preset scene to obtain the matching situation between the current scene represented by the scene feature information and the preset scene, and the corresponding recognition result is obtained. Each preset scene corresponds to one recognition result, and the number of preset scenes is the same as the number of recognition results.

[0070] Step S30: Determine the target scene from the at least two preset scenes based on the at least two recognition results;

[0071] The target scene is the scene that best matches the current scene of the target indoor space from at least two preset scenes. The number of target scenes can be one.

[0072] In one implementation, at least two recognition results are identified as target recognition results, in which the current scene matches the corresponding preset scene or the current scene is the corresponding preset scene. The target scene is then determined based on the preset scene corresponding to the target recognition result.

[0073] In another implementation, the target recognition result is determined by identifying the recognition result in which the probability of the current scene being the corresponding preset scene is greater than the preset probability among at least two recognition results, and the target scene is determined based on the preset scene corresponding to the target recognition result.

[0074] In another implementation, the target recognition result is determined as the recognition result with the highest probability that the current scene is a preset scene among at least two recognition results, and the target scene is determined according to the preset scene corresponding to the target recognition result.

[0075] Step S40: Control at least one smart appliance in the target indoor space to operate according to the target scenario.

[0076] Each preset scene has corresponding preset scene parameters, and different preset scenes correspond to different preset scene parameters. The preset scene parameters include the preset appliances to be turned on and the preset operating parameters of these appliances. The number of preset appliances included in the preset scene parameters can be one or more to accommodate the needs of the corresponding scene. Based on this, the preset appliances in the preset scene parameters corresponding to the target scene are determined as target appliances, and the preset operating parameters corresponding to these appliances are determined as target operating parameters. All target appliances corresponding to the target operating parameters are then controlled to operate according to their respective target operating parameters.

[0077] For example, when the target scenario is a humid weather period, the corresponding target appliance is an air conditioner, which can be controlled to run at the maximum dehumidification level; when the target scenario is a rainy season, the corresponding target appliance is an air conditioner, which can be controlled to run at the maximum dehumidification level; when the target scenario is a sleep scenario, the corresponding target appliance is an air conditioner, which can be controlled to increase the set temperature by a preset amplitude every preset time interval until the total running time of the target scenario reaches the target time; when the target scenario is a dinner party scenario, the corresponding target appliance is an air conditioner, which can be controlled to run at a preset set temperature (e.g., 24 degrees Celsius) and to turn on the fresh air, and so on.

[0078] Furthermore, after step S40, you can return to execute step S10.

[0079] This invention proposes a scene control method for a smart home system. The method sequentially determines the matching status of scene feature information between each of at least two preset scenes and a target indoor space, obtaining at least two recognition results. Based on these two recognition results, a target scene is determined, and at least one smart appliance in the target indoor space is controlled according to the target scene. During this process, the recognition of the target scene comprehensively considers the matching status of all preset scenes, ensuring no unrecognized scenes exist. This effectively improves the accuracy of scene recognition results and guarantees accurate control of smart appliances based on the recognized target scene, thereby effectively improving the accuracy of device control.

[0080] In this embodiment, when the above step S10 is executed during the operation of the smart appliance according to one of at least two preset scenarios, the automatic switching of scenarios can be realized when it is recognized that the target scenario and the smart scenario are operating differently.

[0081] In other embodiments, step S10 may also be performed when the smart appliance is powered on.

[0082] Furthermore, based on the above embodiments, another embodiment of the scene control method for the smart home system of this application is proposed. In this embodiment, reference is made to... Figure 3 Step S20 includes:

[0083] Step S21: Sequentially identify whether the preset scene information of each preset scene in the at least two preset scenes matches the scene feature information, and obtain the at least two identification results.

[0084] The preset scene information specifically refers to the preset conditions that the scene feature information of the corresponding preset scene must meet. The preset scene information may include the target numerical range and / or target information that the scene feature information needs to achieve. The preset scene information can be set by the user according to actual needs, or it can be obtained by the system by analyzing user needs according to certain rules.

[0085] The preset scenario information includes, but is not limited to, at least one of the following: preset environmental information, preset human body information, preset weather information, preset regional information, and preset time information. The preset environmental information includes preset indoor environmental information (e.g., target ranges for indoor temperature, indoor humidity, indoor PM2.5, etc.) and / or preset outdoor environmental information (e.g., target ranges for outdoor temperature, outdoor humidity, outdoor PM2.5, etc.). The preset human body information includes at least one of the following: preset number of people, preset range of personnel location, preset range of respiratory rate, preset range of heart rate, and preset range of body movement. The preset weather information includes sunny, rainy, foggy, or snowy weather. The preset regional information includes preset country / region, preset province / city, preset geographical division (e.g., North, South, North China, Yellow River Basin, etc.), and building climate zone (e.g., hot summer, warm winter, cold). The preset time information includes the target ranges for the year and month, the target ranges for day / night time (e.g., morning or afternoon, day or night, etc.), and the target solar term (e.g., White Dew, Minor Snow, etc.).

[0086] The preset scene information for each preset scene may include preset conditions corresponding to at least two preset dimensions in the corresponding preset scene. The at least two preset dimensions corresponding to different preset scenes may be different. Sub-information corresponding to the scene feature information is extracted according to the at least two preset dimensions corresponding to the preset scene. It is determined whether each sub-information satisfies the preset conditions of the corresponding preset dimension, and the sub-judgment result corresponding to each preset dimension is obtained. The corresponding recognition result is determined based on all sub-judgment results corresponding to all preset dimensions for each preset scene.

[0087] For example, the first preset scenario is the "return to spring" weather scenario, with preset scenario information including: South China, time period February to April, temperature more than 10 degrees Celsius higher than the previous three days, and humidity ≥80%; the second preset scenario is the "plum rain season" scenario, with preset scenario information including: Southern China, time period March to May, and humidity ≥70%. Based on this, when the current scenario feature information includes Guangzhou, April 25th, temperature 28 degrees Celsius, and humidity 86%, the matching result for the first preset scenario is a match between the current scenario and the first preset scenario, and the matching result for the second preset scenario is a match between the current scenario and the second preset scenario.

[0088] In this embodiment, the scene feature information is matched with the preset scene information of each preset scene to ensure that the recognition result can accurately reflect the degree of matching between the current scene represented by the scene feature information and the preset scene, which is conducive to further improving the accuracy of the scene recognition result.

[0089] Furthermore, based on any of the above embodiments, another embodiment of the scene control method for the smart home system of this application is proposed. In this embodiment, reference is made to... Figure 4 Step S20 includes:

[0090] Step S22: Input the scene feature information into the scene recognition model corresponding to each preset scene respectively, and obtain the output result of each scene recognition model. The recognition result includes the output result. The scene recognition model is generated based on machine learning.

[0091] Each preset scenario corresponds to a scene recognition model. The scene recognition model is generated based on machine learning and may include at least one type of machine learning classification model, machine learning regression model, and machine learning clustering model. In this embodiment, the scene recognition models corresponding to different preset scenarios are of the same type. In other embodiments, the scene recognition models corresponding to different preset scenarios may be of different types.

[0092] The output of the scene recognition model includes the probability that the current scene represented by the scene feature information is the corresponding preset scene, or that the current scene represented by the scene feature information is not the corresponding preset scene, or that the current scene represented by the scene feature information is the corresponding preset scene.

[0093] A large amount of scene sample data from indoor spaces is collected before the current moment. The scene sample data is classified according to preset scenarios and assigned corresponding scene labels. Based on the scene labels, positive and negative sample data are determined for each preset scenario. A target model for the corresponding preset scenario is obtained by training a preset machine learning model. The target model can serve as the scene recognition model for the corresponding preset scenario. Positive sample data refers to sample data belonging to the corresponding preset scenario, and negative sample data refers to sample data not belonging to the corresponding preset scenario.

[0094] In this embodiment, the scene recognition model includes a first model or a second model. The first model is used to identify whether the current scene represented by scene feature information is the corresponding preset scene, and the second model is used to identify the probability that the current scene represented by scene feature information is the corresponding preset scene.

[0095] In one implementation of this embodiment, the scene feature information is input into a first model corresponding to each preset scene, and a first output result is obtained for each first model. The first output result includes whether the current scene represented by the scene feature information is the corresponding preset scene. In this embodiment, the first model is a machine learning classification model. In other embodiments, the first model may also be a machine learning model that outputs the same type of result.

[0096] For example, if the preset scenario is a sleep scenario, and the scenario feature information includes: 24 o'clock, respiratory rate of 14 breaths / minute, and heart rate of 75 beats / minute, and the scenario feature information is input into the first model corresponding to the sleep scenario, and the output result of the first model is yes, then the corresponding recognition result is that the current scenario represented by the scenario feature information is a sleep scenario.

[0097] In another implementation of this embodiment, the scene feature information is input into a second model corresponding to each preset scene to obtain a second output result for each second model. The second output result includes the probability that the current scene, represented by the scene feature information, is the corresponding preset scene. In this embodiment, the second model is a machine learning regression model. In other embodiments, the second model may also be a machine learning model that outputs the same type of result.

[0098] For example, if the preset scenario is a dinner party, and the scenario features include: time 11:00 to 14:00, number of people > 8 minutes, and the scenario features are input into the second model corresponding to the dinner party scenario, and the probability that the current scenario represented by the scenario features in the output of the second model is a dinner party scenario is 0.7, then the corresponding recognition result is 0.7.

[0099] In this embodiment, scene feature information is input into the machine learning model of each preset scene for matching, ensuring that the obtained recognition result can accurately reflect the degree of matching between the current scene represented by the scene feature information and the preset scene, which is conducive to further improving the accuracy of scene recognition results.

[0100] Furthermore, in this embodiment, step S22 includes: obtaining the location type of the target indoor space; when the location type is a first type, inputting the scene feature information into a first model corresponding to each preset scene to obtain a first output result corresponding to each first model; when the location type is a second type, inputting the scene feature information into a second model corresponding to each preset scene to obtain a second output result corresponding to each second model; wherein, the first similarity between the at least two preset scenes corresponding to the first type is less than the second similarity between the at least two preset scenes corresponding to the second type.

[0101] In one implementation of this embodiment, the location type is obtained by acquiring parameters input by the user. In another implementation of this embodiment, the location type can be determined by acquiring an image of the target indoor space and determining the location type based on the types of objects placed in the image.

[0102] Location types include living rooms, bedrooms, game rooms, dining rooms, and meeting rooms, etc. Different location types require at least two different preset scenes. The first type corresponds to at least two different preset scenes, and the second type corresponds to at least two different preset scenes.

[0103] For at least two preset scenarios corresponding to the first type, the similarity between each pair of preset scenarios is determined, and a first similarity is determined based on all determined similarities. For at least two preset scenarios corresponding to the second type, the similarity between each pair of preset scenarios is determined, and a second similarity is determined based on all determined similarities.

[0104] In this embodiment, when the similarity between at least two preset scenes corresponding to the target indoor space is small, the first model is used for scene matching analysis, which helps to ensure the accuracy of scene recognition while improving the efficiency of scene recognition; when the similarity between at least two preset scenes corresponding to the target indoor space is large, the second model is used for scene matching analysis, which helps to accurately distinguish similar scenes and further improve the accuracy of scene recognition.

[0105] In other embodiments, one of the first model and the second model can be determined as the scene recognition model according to the user-set parameters, and the scene feature information can be input into the scene recognition model corresponding to each preset scene to obtain the corresponding recognition result.

[0106] Furthermore, in this embodiment, after the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene, the method further includes: when a scene correction instruction input by the user is received, correcting the scene recognition model according to the preset scene corresponding to the scene correction instruction, the scene feature information, and the target scene.

[0107] Furthermore, in this embodiment, after S10, the method further includes: obtaining the operating status parameters of all electrical appliances in the smart home system corresponding to the target indoor space during the scene feature information detection process, correcting the target model trained based on the scene sample data according to the operating status parameters and the preset scene, and obtaining the scene recognition model corresponding to the preset scene.

[0108] Furthermore, based on any of the above embodiments, another embodiment of the scene control method for the smart home system of this application is proposed. In this embodiment, step S20 includes: when there are user setting parameters corresponding to the preset scene, determining the preset scene information corresponding to the preset scene according to the user setting parameters, and executing step S21; when there are no user setting parameters corresponding to the preset scene, executing step S22.

[0109] For preset scenarios with corresponding user-defined parameters, the corresponding preset scenario information can be determined according to the user-defined parameters. For preset scenarios without corresponding user-defined parameters, the pre-stored scenario information in the memory can be used as the corresponding preset scenario information. In other words, the preset scenario information corresponding to each preset scenario can be determined according to the user-defined parameters, or the preset scenario information corresponding to some preset scenarios can be set by the user, while the preset scenario information corresponding to other preset scenarios can be pre-stored parameters.

[0110] In this embodiment, the scene recognition accuracy of the scene recognition model generated by machine learning is higher than that of the scene recognition using preset scene information. Therefore, when the user does not set the scene himself, using the machine learning model for scene recognition is beneficial to improving the accuracy of the scene recognition results. However, when the user sets the scene himself, the pre-trained machine learning model may deviate from the user's needs. In this case, performing scene recognition according to the preset scene information set by the user is beneficial to ensure that the target scene obtained by scene recognition can accurately match the user's needs, thereby further improving the accuracy of scene recognition.

[0111] Furthermore, based on any of the above embodiments, another optional embodiment of the scene control method for the smart home system of this application is proposed. In this embodiment, reference is made to... Figure 5 Step S30 includes:

[0112] Step S31: When there is more than one recognition result that satisfies the preset conditions among the at least two recognition results, the target scene is determined according to the priority or prediction probability of all the preset scenes corresponding to all recognition results that satisfy the preset conditions.

[0113] Wherein, the preset condition indicates that the preset scene matches the scene feature information, and the prediction probability is the probability that the current scene represented by the scene feature information is the corresponding preset scene.

[0114] In one implementation of this embodiment, the recognition result includes whether the current scene represented by the scene feature information is a corresponding preset scene, and the preset condition includes the recognition result being that the current scene represented by the scene feature information is a corresponding preset scene. Here, the recognition result can be the recognition result obtained from the analysis in step S21 above or the recognition result obtained based on the analysis of the first model above.

[0115] In another implementation of this embodiment, the recognition result includes the predicted probability, and the preset condition includes the predicted probability being greater than the first preset probability. Here, the recognition result can be the recognition result obtained from the analysis in step S22 above.

[0116] In this embodiment, the preset scene corresponding to the recognition result that meets the preset conditions is the candidate scene, the preset scene with the highest priority among all candidate scenes is the target scene, and the preset scene with the highest probability among all candidate scenes is the target scene.

[0117] The priority of each candidate scenario can be a pre-set fixed priority or it can vary according to the actual situation. For example, the priority of each candidate scenario can be determined based on the different combination characteristics of all the determined candidate scenarios; different scenarios included in the determined candidate scenarios will result in different priorities for the candidate scenarios. Another example is that the priority of each candidate scenario can be determined based on the types of all smart appliances that are not currently turned on in the smart home system.

[0118] The predicted probability corresponding to the alternative scenarios can be obtained by inputting the scenario feature information into the second model mentioned above. When the recognition result includes the predicted probability, the predicted probability in the device result can be used directly.

[0119] In this embodiment, when there are more than one preset scene and scene feature information matching based on at least two recognition results, the target scene for scene control of the smart home system is determined based on the priority or predicted probability corresponding to the preset scene that meets the preset conditions. This ensures that when the smart appliances are controlled by the determined target scene, the environmental state of the target indoor space can be accurately matched with the user's needs, thereby further improving the accuracy of scene recognition and device control.

[0120] In other embodiments, all candidate scenarios can be used as target scenarios, and preset scenario parameters corresponding to more than one target scenario can be obtained. The set of preset electrical appliances included in the more than one preset scenario parameters can be used as the target electrical appliance set to be turned on. When there is a first electrical appliance in the target electrical appliance set, the first electrical appliance is an electrical appliance with different preset operating parameters in different target scenarios. Then, the weight value of the preset scenario parameter corresponding to each target scenario is determined according to the priority or predicted probability corresponding to each target scenario. The weighted average of all preset operating parameters corresponding to the first electrical appliance is calculated according to the weight value to obtain the target operating parameters corresponding to the first electrical appliance. When there is a second electrical appliance in the target electrical appliance set, the second electrical appliance has a corresponding preset operating parameter in a single target scenario or has the same preset operating parameter in different target scenarios. Then, the preset operating parameter corresponding to the second electrical appliance is used as the target operating parameter corresponding to the second electrical appliance.

[0121] Furthermore, based on any of the above embodiments, another optional embodiment of the scene control method for the smart home system of this application is proposed. In this embodiment, after step S30, the method further includes: outputting switching prompt information corresponding to the target scene; when receiving confirmation information corresponding to the target scene, executing the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene.

[0122] The switching prompt message can be output in at least one of the following ways: text, sound, vibration, light, etc.

[0123] The switching prompt message is used to prompt the user to confirm whether to switch the current operating scene of the smart appliance to the target scene.

[0124] The confirmation information is entered by the user based on the switching prompt.

[0125] In addition, when a cancellation message corresponding to the target scenario is received or no confirmation message corresponding to the target scenario is received within a preset time period, the smart appliance can be controlled to maintain its current state of operation.

[0126] In this embodiment, after determining the target scene, a switching prompt message is sent to the user. Upon receiving confirmation from the user, the smart appliance is switched to operate in the target scene. This helps to ensure the accuracy of the smart appliance's operation and control, and ensures that the state of the target indoor space can accurately meet the user's needs.

[0127] In other embodiments, step S40 can be executed directly after step S30.

[0128] Furthermore, in this embodiment, after step S30, the method further includes: when the predicted probability corresponding to the target scene is greater than the second preset probability, executing the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene; when the predicted probability corresponding to the target scene is less than or equal to the second preset probability, outputting the switching prompt information corresponding to the target scene; and when receiving the confirmation information corresponding to the target scene, executing the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene; wherein, the predicted probability is the probability that the current scene represented by the scene feature information is the target scene.

[0129] The second preset probability here is greater than the first preset probability mentioned above.

[0130] In this embodiment, when the current scenario is highly likely to be the target scenario, the smart appliance directly switches to the target scenario; when the current scenario is less likely to be the target scenario, a switching prompt message is output, and the smart appliance switches to the target scenario only after receiving the confirmation message. This helps to ensure the accuracy of the smart appliance's scenario switching and ensures that user needs are accurately met.

[0131] Furthermore, this invention also proposes a storage medium storing a scene control program for a smart home system. When the scene control program for the smart home system is executed by a processor, it implements the relevant steps of any of the above embodiments of the scene control method for the smart home system.

[0132] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0133] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0134] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, smart appliance, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0135] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A scene control method for a smart home system, characterized in that, The scene control method of the smart home system includes the following steps: Obtain scene feature information of the target indoor space; Sequentially determine the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results; The target scene in the at least two preset scenarios is determined based on the at least two recognition results; Control the operation of at least one smart appliance in the target indoor space according to the target scenario; The step of determining the target scene among the at least two preset scenes based on the at least two recognition results includes: When there is more than one recognition result that satisfies the preset conditions among the at least two recognition results, the target scene is determined according to the priority or prediction probability of all the preset scenes corresponding to all recognition results that satisfy the preset conditions, and the prediction probability is the probability that the current scene represented by the scene feature information is the corresponding preset scene. The identification result includes whether the current scene represented by the scene feature information is the corresponding preset scene, and the preset condition includes whether the identification result is that the current scene represented by the scene feature information is the corresponding preset scene; or, the identification result includes the prediction probability, and the preset condition includes that the prediction probability is greater than a first preset probability.

2. The scene control method for a smart home system as described in claim 1, characterized in that, The step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes: Sequentially identify whether the preset scene information of each of the at least two preset scenarios matches the scene feature information, and obtain the at least two identification results.

3. The scene control method for a smart home system as described in claim 2, characterized in that, The step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes: When there are user setting parameters corresponding to the preset scenario, the preset scenario information corresponding to the preset scenario is determined according to the user setting parameters, and the preset scenario information of each preset scenario in the at least two preset scenarios is sequentially identified to see if it matches the scenario feature information, so as to obtain the at least two identification results; When no user setting parameters corresponding to the preset scenario exist, the scenario feature information is input into the scenario recognition model corresponding to each preset scenario to obtain the output result of each scenario recognition model. The recognition result includes the output result. The scenario recognition model is generated based on machine learning.

4. The scene control method for a smart home system as described in claim 1, characterized in that, The step of sequentially determining the matching situation between various preset scenarios and the scene feature information in at least two preset scenarios to obtain at least two recognition results includes: The scene feature information is input into the scene recognition model corresponding to each preset scene, and the output result of each scene recognition model is obtained. The recognition result includes the output result. The scene recognition model is generated based on machine learning.

5. The scene control method for a smart home system as described in claim 4, characterized in that, The scene recognition model includes a first model or a second model. The step of inputting the scene feature information into the scene recognition model corresponding to each preset scene and obtaining the output result corresponding to each scene recognition model includes: The scene feature information is input into the first model corresponding to each preset scene, and a first output result is obtained for each first model. The first output result includes whether the current scene represented by the scene feature information is the corresponding preset scene. Alternatively, the scene feature information can be input into the second model corresponding to each preset scene to obtain a second output result corresponding to each second model. The second output result includes the probability that the current scene represented by the scene feature information is the corresponding preset scene.

6. The scene control method for a smart home system as described in claim 5, characterized in that, The step of inputting the scene feature information into the scene recognition model corresponding to each preset scene and obtaining the output result of each scene recognition model includes: Obtain the location type of the target indoor space; When the location type is the first type, the scene feature information is input into the first model corresponding to each preset scene to obtain the first output result corresponding to each first model; When the location type is the second type, the scene feature information is input into the second model corresponding to each preset scene to obtain the second output result corresponding to each second model; Wherein, the first similarity between the at least two preset scenarios corresponding to the first type is less than the second similarity between the at least two preset scenarios corresponding to the second type.

7. The scene control method for a smart home system as described in any one of claims 1 to 6, characterized in that, After the step of determining the target scene among the at least two preset scenes based on the at least two recognition results, the method further includes: Output the switching prompt information corresponding to the target scene; When a confirmation message corresponding to the target scenario is received, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scenario is executed.

8. The scene control method for a smart home system as described in claim 7, characterized in that, After the step of determining the target scene among the at least two preset scenes based on the at least two recognition results, the method further includes: When the predicted probability corresponding to the target scenario is greater than the second preset probability, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scenario is executed. When the predicted probability corresponding to the target scene is less than or equal to the second preset probability, the switching prompt information corresponding to the target scene is output. When the confirmation information corresponding to the target scene is received, the step of controlling the operation of at least one smart appliance in the target indoor space according to the target scene is executed.

9. A scene control device for a smart home system, characterized in that, The scene control device of the smart home system includes: a memory, a processor, and a scene control program of the smart home system stored in the memory and executable on the processor. When the scene control program of the smart home system is executed by the processor, it implements the steps of the scene control method of the smart home system as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium stores a scene control program for a smart home system. When the scene control program for the smart home system is executed by the processor, it implements the steps of the scene control method for the smart home system as described in any one of claims 1 to 8.

Images