Intelligent control method, device, medium and program product based on large model

Abstract

The application provides a large model-based intelligent control method, device, medium and program product, and relates to the technical field of smart home / smart home. The method comprises the following steps: inputting a parameter value of a target environment parameter, position information of at least part of intelligent household appliance devices in a room and a first prompt word into an analysis large model to obtain an analysis result, wherein the first prompt word is used to instruct the analysis large model to analyze the indoor distribution and / or change trend of the target environment parameter according to the parameter value and the position information; inputting the analysis result and a second prompt word into a control large model to obtain a control result, wherein the second prompt word is used to instruct the control large model to generate a control instruction for at least one target intelligent household appliance device in a household appliance set according to the analysis result, and the control instruction is used to control the target intelligent household appliance device to adjust the target environment parameter. The method provided in the application can provide appropriate control for the intelligent household appliance devices and improve user experience.

Description

Technical Field

[0001] This application relates to the field of smart home / intelligent home technology, and in particular to a smart control method, device, medium and program product based on a large model. Background Technology

[0002] IoT cloud is a technology that combines the Internet of Things (IoT) with cloud computing. The IoT connects various physical devices through the Internet to collect, transmit, and process data; cloud computing provides powerful computing and storage capabilities to process and analyze large amounts of IoT data.

[0003] With the development of intelligent home appliances, users' requirements for the intelligence of home appliances are gradually increasing. In the existing technology, the control logic of some different home appliances is set in advance to control the home appliances in a coordinated manner, or the home appliances are controlled by models. Usually, the control of home appliances is based on limited data, which makes it difficult to provide appropriate control for the actual use of home appliances by users, thus affecting the user experience.

[0004] Therefore, this application proposes an intelligent control method based on a large model to solve the above problems. Summary of the Invention

[0005] This application provides a large-scale intelligent control method, device, medium, and program product to solve the problem in the prior art that it is difficult to provide suitable device control for the actual situation of users.

[0006] Firstly, this application provides a large-scale model-based intelligent control method for controlling a collection of home appliances, the collection including multiple intelligent home appliances, the intelligent home appliances being used to collect and / or adjust environmental parameters, the method comprising:

[0007] Obtain the parameter values ​​of the target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter;

[0008] The parameter values, the indoor location information of at least some smart home appliances, and the first prompt word are input into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameters based on the parameter values ​​and location information.

[0009] The analysis results and the second prompt word are input into the control model to obtain control results. The second prompt word is used to instruct the control model to generate control instructions for at least one target smart home appliance in the set of home appliances based on the analysis results. The control instructions are used to control the target smart home appliance to adjust the target environmental parameters.

[0010] In one possible implementation, the parameter values, the location information of at least some of the smart home appliances indoors, and the first prompt word are input into the large-scale analysis model to obtain analysis results, including:

[0011] Every preset period, input the following information and the first prompt word into the large analysis model to obtain the analysis results for the current period:

[0012] The current parameter value and historical parameter value corresponding to the current period;

[0013] The location information of at least some of the smart home appliances indoors;

[0014] Analysis results from the previous period;

[0015] The target parameter values ​​corresponding to the target environmental parameters;

[0016] Specifically, the first prompt word is used to indicate: based on the parameter values ​​and location information of the current period, combined with the target parameter values ​​and the analysis results of the previous period, analyze the indoor distribution and / or trend of the target environmental parameters in the current period.

[0017] In one possible implementation, the analysis results and the second prompt word are input into the large control model to obtain control results, including:

[0018] Every preset period, the following information and the second prompt word are input into the control model to obtain the control result for the current period:

[0019] Analysis results corresponding to the current cycle;

[0020] The control command corresponding to the previous cycle;

[0021] The target parameter values ​​corresponding to the target environmental parameters;

[0022] The second prompt word is specifically used to indicate: based on the analysis results of the current cycle, combined with the target parameter value and the control command of the previous cycle, the control result of the current cycle is obtained;

[0023] The target parameter value can be set in any of the following ways: default setting for smart home appliances, or user setting for smart home appliances.

[0024] In one possible implementation, the information input to the control model further includes:

[0025] The list of smart home appliances in the home appliance collection;

[0026] Each smart home appliance has its own set of functions; each function in the set is used to adjust one or more environmental parameters.

[0027] The second prompt word is also used to indicate: based on the function sets corresponding to each smart home appliance, at least one target smart home appliance is determined from the set of appliances to adjust the target environmental parameters.

[0028] In one possible implementation, the method further includes:

[0029] Obtain a training sample set and train the large control model based on the training sample set;

[0030] The training samples in the training sample set include analysis result samples, control command samples from the previous cycle, target parameter value samples, and control command samples from the current cycle.

[0031] When the analysis result sample represents the target environmental parameter approaching the target parameter value sample at a preset speed, the control command sample of the current cycle is the same as the control command sample of the previous cycle.

[0032] In one possible implementation, the parameter values ​​are indoor parameter values; the information input to the control model also includes at least one of the following:

[0033] The target environmental parameter is the outdoor parameter value, wherein the outdoor parameter value is obtained by querying the geographical location information of the appliance set.

[0034] User preference information includes target parameter values, the speed at which the target environmental parameters approach the target parameter values, and the usage time periods of the respective functional groups of each smart home appliance.

[0035] Secondly, this application provides a large-scale intelligent control device for controlling a collection of home appliances, the collection including multiple intelligent home appliances, the intelligent home appliances being used to collect and / or adjust environmental parameters, the device comprising:

[0036] The acquisition module is used to acquire the parameter values ​​of target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter;

[0037] The first processing module is used to input the parameter values, the location information of at least some smart home appliances in the room, and the first prompt word into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameters based on the parameter values ​​and location information.

[0038] The second processing module is used to input the analysis results and the second prompt word into the control model to obtain the control result. The second prompt word is used to instruct the control model to generate a control command for at least one target smart home appliance in the set of home appliances based on the analysis results. The control command is used to control the target smart home appliance to adjust the target environmental parameters.

[0039] Thirdly, this application provides an electronic device, including: at least one processor and a memory;

[0040] The memory stores computer-executed instructions;

[0041] The at least one processor executes the computer execution instructions stored in the memory, causing the at least one processor to perform the large-model-based intelligent control method as described above.

[0042] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the intelligent control method based on a large model as described above.

[0043] Fifthly, this application provides a computer program product, the computer program product including instructions, which, when executed on an electronic device, cause the electronic device to implement the intelligent control method based on a large model as described above.

[0044] This application provides a large-scale model-based intelligent control method, device, medium, and program product for controlling a set of home appliances, including multiple intelligent home appliances. These intelligent home appliances are used to collect and / or adjust environmental parameters. The method includes: acquiring parameter values ​​of target environmental parameters collected by at least some of the intelligent home appliances in the set; the target environmental parameter is any environmental parameter that can be collected by the set of home appliances, and the parameter values ​​include historical and / or current parameter values ​​of the target environmental parameter; inputting the parameter values, the indoor location information of the at least some intelligent home appliances, and a first prompt word into an analysis large-scale model to obtain analysis results, wherein the first prompt word instructs the analysis large-scale model to analyze the indoor distribution and / or trend of the target environmental parameter based on the parameter values ​​and location information; inputting the analysis results and a second prompt word into a control large-scale model to obtain control results, wherein the second prompt word instructs the control large-scale model to generate control instructions for at least one target intelligent home appliance in the set based on the analysis results, and the control instructions are used to control the target intelligent home appliance to adjust the target environmental parameter.

[0045] In the above method, a large model is used to analyze and control various types of information from multiple smart home appliances in the appliance set. For any environmental parameter, as the target environmental parameter, at least some smart home appliances in the appliance set can collect parameter values. The collected parameter values, the location information of at least some smart home appliances, and the first prompt word can be used as input to the analysis large model. The first prompt word instructs the analysis large model to analyze the indoor distribution and / or changing trend of the target environmental parameter based on the input, and obtain the analysis results. The analysis results and the second prompt word are input into the control large model. The second prompt word instructs the control large model to generate at least one target smart home appliance in the appliance set based on the input, and the target smart home appliance adjusts according to the control command. With reasonable information as input to the large model, the appliance set is reasonably controlled, so that the smart home appliances create a suitable environment for the user and meet the user experience. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 A hardware environment diagram for intelligent control based on a large model is provided for embodiments of this application;

[0048] Figure 2 System framework diagram of an intelligent control based on a large model provided by an embodiment of the present application;

[0049] Figure 3 Flow schematic of an intelligent control method based on a large model provided by an embodiment of the present application Figure 1 ;

[0050] Figure 4 Flow schematic of an intelligent control method based on a large model provided by an embodiment of the present application Figure 2 ;

[0051] Figure 5 Diagram of an intelligent control device based on a large model provided by an embodiment of the present invention;

[0052] Figure 6 Hardware schematic diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0053] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0054] It should be noted that the terms "first", "second", etc. in the present application are used to distinguish similar objects and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order different from those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0055] Intelligent household appliances are gradually popularized among广大用户, and with the intelligent development of household appliances, users' requirements for the intelligence of household appliances are gradually increasing; in the prior art, the control logics of some different household appliances are set in advance in a room to perform interlocking control on the household appliances, or the household appliances are controlled through a model, usually controlling a single household appliance by referring to limited data, which is difficult to provide appropriate control according to the actual usage situation of the household appliances by users and affects the user experience.

[0056] Therefore, this application proposes a large-model-based intelligent control method that provides control of smart home appliances based on diversified information analysis to enhance user experience.

[0057] The implementation process of the intelligent control method based on a large model proposed in this application is described below with reference to the accompanying drawings and specific embodiments.

[0058] Figure 1 This application provides a hardware environment diagram for intelligent control based on a large model; according to one aspect of this application, an intelligent control method based on a large model is provided. This intelligent control method based on a large model is widely used in whole-house intelligent digital control application scenarios such as smart homes, smart home ecosystems, and intelligence house ecosystems. Optionally, in this embodiment, the above method can be applied to, for example... Figure 1 The hardware environment shown consists of terminal device 102 and server 104. For example... Figure 1 As shown, server 104 is connected to terminal device 102 via a network and can be used to provide services (such as application services) to the terminal or clients installed on the terminal. A database can be set up on the server or independently of the server to provide data storage services for server 104. Cloud computing and / or edge computing services can be configured on the server or independently of the server to provide data processing services for server 104.

[0059] The aforementioned network may include, but is not limited to, at least one of the following: wired network, wireless network. The aforementioned wired network may include, but is not limited to, at least one of the following: wide area network, metropolitan area network, local area network. The aforementioned wireless network may include, but is not limited to, at least one of the following: Wi-Fi (Wireless Fidelity), Bluetooth. The terminal device 102 may not be limited to PC, mobile phone, tablet computer, smart air conditioner, smart range hood, smart refrigerator, smart oven, smart stove, smart washing machine, smart water heater, smart washing equipment, smart dishwasher, smart projector, smart TV, smart clothes rack, smart curtains, smart audio-visual equipment, smart socket, smart speaker, smart speaker box, smart fresh air equipment, smart kitchen and bathroom equipment, smart bathroom equipment, smart robot vacuum cleaner, smart window cleaning robot, smart mopping robot, smart air purifier, smart steam oven, smart microwave oven, smart water heater, smart air purifier, smart water dispenser, smart door lock, etc.

[0060] Figure 2 This document presents a system framework diagram for intelligent control based on a large model, as provided in an embodiment of this application. Figure 2As shown, the system includes: terminals, smart home appliances, routers, and the cloud; the terminals can communicate with the cloud through protocols, gateways, etc.; the smart home appliances can communicate with the cloud through the router and then through the gateway.

[0061] Smart home appliances include air conditioning devices such as dehumidifiers, air conditioners, and thermometers. These devices can create a comfortable room environment for users by adjusting temperature, humidity, formaldehyde concentration, and indoor PM2.5 concentration.

[0062] Smart home appliances can transmit current and historical data related to the device to the cloud, such as the ambient temperature collected by the device; users can also transmit preference data to the cloud through their terminals, allowing the cloud to determine the target data based on user preferences. For example, if a user prefers a warm environment, the target data set by the cloud can be biased towards warmth.

[0063] The cloud platform includes a gateway, data processing service, data computing service, large-scale analysis model, and large-scale control model. The gateway is responsible for receiving data from smart home appliances and terminals. The data processing service is responsible for storing data and triggering calculations. The data computing service is responsible for organizing the data required by different large-scale analysis models. The large-scale analysis models are responsible for preliminary evaluation. The large-scale control model is responsible for analyzing how to accurately control the corresponding home appliances based on the evaluation results of the large-scale analysis models and the data of the home appliances to be controlled, so as to provide users with a suitable environment and seamlessly improve the user experience.

[0064] Data such as the capabilities of smart home appliances, the types of different attributes, the range of values, and the applicable scenarios, as well as data such as user habits and preferences, are stored in cloud storage in the form of vector data. Data processing services include cloud storage.

[0065] Developers can provide accurate, adjusted, checked, and validated data to fine-tune the large analytical model using LoRA.

[0066] Developers provide the initial data to the control model and pre-train it. Subsequently, LoRA can be used to optimize the control model by combining data such as user habits and preferences.

[0067] Users can add their own preference data on the mobile app, such as their preference for a warm environment; and set upper and lower thresholds for the data, such as the range of temperature and humidity control.

[0068] Indoor air quality data reported by smart air devices in the user's home. Smart air devices include, but are not limited to, air conditioners, fresh air systems, smart thermometers and hygrometers, and humidifiers / dehumidifiers; the reported air data includes, but is not limited to, temperature, humidity, formaldehyde concentration, and indoor PM2.5 concentration.

[0069] Indoor air quality data reported by the device, user control data of the device, and user preference settings are stored in cloud storage in the form of vector data.

[0070] The indoor air condition data reported by the gateway is processed by the data computing service to obtain an accurate indoor condition, which is then saved to cloud storage. At the same time, the latest indoor air data is provided to the control model.

[0071] After receiving indoor air quality data, the control model calculates whether to adjust the indoor air quality by controlling the air quality devices in the user's home (e.g., adjusting the power / fan speed of the fresh air system if high formaldehyde concentration is detected). If control is required, the devices are controlled through the gateway, and the calculation results of the control model are saved to the cloud database.

[0072] Figure 3 A flowchart illustrating a large-model-based intelligent control method provided in this application embodiment. Figure 1 .like Figure 3 As shown, a method for controlling a collection of home appliances, including multiple smart home appliances, wherein the smart home appliances are used to collect and / or adjust environmental parameters, includes:

[0073] S301. Obtain the parameter values ​​of the target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter.

[0074] Environmental parameters can be adjusted by smart home appliances to reach a suitable state to meet user needs; analyzing historical and / or current parameter values ​​facilitates the adjustment of future parameter values; there are various types of environmental parameters, each type corresponding to a large analysis model, so one large analysis model is responsible for one target environmental parameter; environmental parameters include temperature, humidity, formaldehyde concentration, indoor PM2.5 concentration, etc.

[0075] A home appliance collection can be a group of multiple smart home appliances in a room of a household; some smart home appliances have the function of collecting environmental parameters, some have the function of controlling environmental parameters, and some have the function of collecting and controlling environmental parameters; smart home appliances are smart home appliances connected to the cloud, and smart home appliances have the function of collecting data and reporting data to the cloud; for example, a thermometer can only collect temperature, an old-fashioned heater can only adjust temperature, and a new-style heater and air conditioner can both collect temperature and adjust temperature;

[0076] Some smart home appliances in the appliance collection acquire parameter values ​​related to the target environment and send them to the cloud data processing service through the gateway. The data processing service can store the data and also send the data to the data computing service; the data computing service sends the data to the large analysis model.

[0077] S302. Input the parameter values, the location information of at least some smart home appliances in the room, and the first prompt word into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameters based on the parameter values ​​and location information.

[0078] The large-scale analysis model is trained in advance using input samples. Parameter values ​​and the location information of some smart home appliances indoors can be input into the corresponding large-scale analysis model. The large-scale analysis model performs preliminary analysis based on the first prompt word, and the analysis result can be a sentence of text; for example, if the target environmental parameter is temperature, the evaluation result is that the temperature tends to rise. The first prompt word provides instructions or frameworks for the large-scale analysis model to output. The content of the first prompt word is such as: "You are an indoor air analysis expert. Combining parameter values ​​and location information, analyze the target environmental parameters," and corresponding instructions.

[0079] Data computing services can classify data according to target environmental parameters and input it into the corresponding large-scale analysis model. The large-scale analysis model integrates parameter values ​​and location information to analyze the high and low distribution of target environmental parameters at location. For example, smart home appliances A and B are located in different locations in a room (such as the center of the room / near a window in the room), and the temperature obtained will be different. The high and low comparison value may be that the temperature at A is higher than that at B.

[0080] This allows for a quick determination of the target environmental parameters, facilitating their input into subsequent large-scale control model calculations.

[0081] The data computing service assigns temperature-related data, humidity-related data, and other data to their respective matching large-scale analysis models for evaluation; the various large-scale analysis models can perform parallel calculations on the data they receive, accelerating the initial evaluation.

[0082] S303. Input the analysis results and the second prompt word into the control model to obtain the control results. The second prompt word is used to instruct the control model to generate control instructions for at least one target smart home appliance in the set of home appliances based on the analysis results. The control instructions are used to control the target smart home appliance to adjust the target environmental parameters.

[0083] The target smart home appliance is a smart home appliance connected to the cloud and capable of adjusting target environmental parameters. The control model is pre-trained using input samples. The analysis results of the analysis model are input into the control model. Based on the second prompt word and the analysis results, the control model can select at least one target smart home appliance from the set of appliances for control and generate corresponding control commands to control the target smart home appliance for adjustment. For example, if the analysis results indicate that the room temperature is rising slowly and the user desires a warm environment, the smart home appliance needs to reach the warm environment at a certain speed. In this case, the control command output by the control model can control the air conditioner to increase the airflow to quickly reach the target temperature.

[0084] In this embodiment, a large model is used to analyze and control various types of information from multiple smart home appliances in a set of home appliances. For any environmental parameter, as a target environmental parameter, at least some smart home appliances in the set of home appliances can collect parameter values. The collected parameter values, the location information of at least some smart home appliances, and a first prompt word can be used as input to the large analysis model. The first prompt word instructs the large analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameter based on the input, and obtain the analysis results. The analysis results and the second prompt word are input to the control model. The second prompt word instructs the control model to generate at least one target smart home appliance in the set of home appliances based on the input, and the target smart home appliance adjusts according to the control command. By using reasonable information as input to the large model, the set of home appliances is reasonably controlled, so that the smart home appliances create a suitable environment for the user and meet the user experience.

[0085] In one embodiment, analyzing a large model may include more input information:

[0086] For example, the parameter values, the location information of at least some of the smart home appliances indoors, and the first prompt word are input into the large-scale analysis model to obtain the analysis results, including:

[0087] Every preset period, input the following information and the first prompt word into the large analysis model to obtain the analysis results for the current period:

[0088] The current parameter value and historical parameter value corresponding to the current period;

[0089] The location information of at least some of the smart home appliances indoors;

[0090] Analysis results from the previous period;

[0091] The target parameter values ​​corresponding to the target environmental parameters;

[0092] Specifically, the first prompt word is used to indicate: based on the parameter values ​​and location information of the current period, combined with the target parameter values ​​and the analysis results of the previous period, analyze the indoor distribution and / or trend of the target environmental parameters in the current period.

[0093] The analysis model and the control model work together, and both can repeatedly analyze and control environmental parameters within the same cycle, which can be set to 30 minutes.

[0094] The current and historical parameter values ​​have corresponding acquisition periods, which are shorter than the working period of the large model; for example, the acquisition period is 10 seconds or 1 minute.

[0095] Over the past 30 minutes, smart home appliances have collected target environmental parameters every minute. The last collected parameter value is the current parameter value, and the parameters collected before the last time are historical parameter values. Each smart home appliance corresponds to a set of parameter values, which can be grouped and input into the large analysis model. Each set of parameter values ​​can also be sorted according to the historical collection time points, and the parameter values ​​collected by each smart home appliance at each time point can be integrated separately. Integration methods include averaging, weighted summation, etc. For example, assuming that the air conditioner collects a temperature of 22℃ and the thermometer collects a temperature of 22.3℃ at 11:00, and the air conditioner collects a temperature of 22.1℃ and the thermometer collects a temperature of 22.5℃ at 11:01, with the same weight, then the integrated parameter values ​​at 11:00 and 11:01 are 22.05℃ and 22.4℃, respectively. The integrated parameter values ​​can also be used as input to the large analysis model.

[0096] The analysis model inputs at least some of the smart home appliances' location information indoors, which facilitates the comparison of location information and parameter values ​​between the large model and the analysis of the distribution of parameter values ​​in different locations in the room; some location information corresponds to high parameter values, while other location information corresponds to low parameter values;

[0097] Analyzing the target environmental parameters input to the large model helps to confirm whether the target environmental parameters are currently approaching the target parameter values.

[0098] By inputting the analysis results from the previous period into the large analysis model, the large analysis model can gain an understanding of the analysis results from the previous period. It can also combine the indoor distribution and / or trend of the target environmental parameters in the previous period to confirm the indoor distribution and / or trend of the target environmental parameters of the smart home appliances in the current period.

[0099] When there is more input information, the first prompt word suggests that the large analysis model obtains analysis results based on the input information, obtains the indoor distribution and / or trend of the target environmental parameters, and repeatedly obtains the current analysis results according to the cycle.

[0100] In one embodiment, controlling the large model may include more input information:

[0101] For example, the analysis results and the second prompt word are input into the large control model to obtain the control results, including:

[0102] Every preset period, the following information and the second prompt word are input into the control model to obtain the control result for the current period:

[0103] Analysis results corresponding to the current cycle;

[0104] The control command corresponding to the previous cycle;

[0105] The target parameter values ​​corresponding to the target environmental parameters;

[0106] The second prompt word is specifically used to indicate: based on the analysis results of the current cycle, combined with the target parameter value and the control command of the previous cycle, the control result of the current cycle is obtained;

[0107] The target parameter value can be set in any of the following ways: default setting for smart home appliances, or user setting for smart home appliances.

[0108] The control model can integrate multiple inputs to obtain control results; the analysis results can be output from the analysis model and sent to the control model; other inputs to the control model can be sent to the control model through gateways and data processing services; the analysis results corresponding to the current period help the control model to roughly understand the environmental situation; the control commands corresponding to the previous period allow the control model to know the past control situation; the target parameter values ​​corresponding to the target environmental parameters allow the control model to know the environmental conditions that should be achieved; based on these inputs, the control model can obtain the control situation for the next period.

[0109] For example, the information input to the control model also includes:

[0110] The list of smart home appliances in the home appliance collection;

[0111] Each smart home appliance has its own set of functions; each function in the set is used to adjust one or more environmental parameters.

[0112] The second prompt word is also used to indicate: based on the function sets corresponding to each smart home appliance, at least one target smart home appliance is determined from the set of appliances to adjust the target environmental parameters.

[0113] The control model knows which devices it can control based on the list of smart home appliances in the appliance collection; each smart home appliance has its corresponding function set, and adjusting the working state of each function in the function set can affect the target environmental parameters, thereby adjusting the target environmental parameters to change; adjustments are made in conjunction with functions, for example, the temperature range that an air conditioner can adjust is 16 degrees Celsius to 30 degrees Celsius, and what fan speeds it has; a heater can adjust the temperature.

[0114] For example, the parameter values ​​are indoor parameter values; the information input to the control model also includes at least one of the following:

[0115] The target environmental parameter is the outdoor parameter value, wherein the outdoor parameter value is obtained by querying the geographical location information of the appliance set.

[0116] User preference information includes target parameter values, the speed at which the target environmental parameters approach the target parameter values, and the usage time periods of the respective functional groups of each smart home appliance.

[0117] The location information of smart home appliances is indoor information; the control model can also refer to outdoor parameter values. Different users in different geographical locations have different external climates and dryness levels, which can be used as a control reference for adjusting the environment.

[0118] To make the control results more user-friendly and provide a comfortable experience, user preference information can be used as input to the control model. For example, if a user prefers a warm environment and low noise at night, the control model can control smart home appliances to output higher temperatures within a suitable range and to disable the high-noise heating function at night.

[0119] Figure 4 A flowchart illustrating an intelligent control method based on a large model provided in this embodiment of the invention. Figure 2 ,like Figure 4 As shown, smart home appliances collect relevant data on target environmental parameters and send them to a data processing service for storage via a gateway. The data computing service can obtain data from the data processing service and organize the data required for different target environmental parameters. Different target environmental parameters can then be assigned to different large-scale analysis models. The large-scale analysis model and the large-scale control model obtain analysis results and control results based on the input. Both analysis results and control results can be stored in the data processing service. The control commands in the control results are sent to the target smart home appliances via the gateway.

[0120] Table 1 shows the input and output diagrams of the large-scale analysis model; Table 2 shows the input and output diagrams of the large-scale control model.

[0121] Table 1

[0122]

[0123]

[0124] Table 2

[0125]

[0126] As shown in Tables 1 and 2, taking temperature as an example of the target environmental parameter, smart home appliances include air conditioners, thermometers, and heaters.

[0127] Analysis of large model inputs:

[0128] Equipment list: Air conditioner, thermometer, heater;

[0129] Equipment locations: Air conditioner is near the window in the room, thermometer is in the center of the room, and heater is near the door in the room; some heaters have temperature collection functions, while others do not; if the heater has temperature collection functions, the analysis model input does not include the information that the heater is near the door in the room.

[0130] Temperature values: Three sets of temperature values ​​from the air conditioner, thermometer, and heater within a period of time. Each set of temperature values ​​includes multiple historical temperature values ​​and a current temperature value; a combined set of temperature values ​​at a given time point (historical temperature curve, which may not be continuous and can be a broken line); n is an integer number.

[0131] Target temperature value: This can be a specific value (25℃), a temperature range (24℃-26℃), or a descriptive word (warm).

[0132] Controlling the input of the large model:

[0133] Analyze the results of the large model;

[0134] Equipment list: Air conditioner, heater;

[0135] Equipment location: Air conditioner near the window in the room; heater near the door in the room.

[0136] Function set: Air conditioner temperature setting, fan speed level, air direction setting; Heater temperature setting;

[0137] User preferences: Long-term air conditioning optimization (hoping it won't reach the target temperature too quickly), night control (allowing for noisier functions to be used at night), and a preference for warmth;

[0138] Historical control: The air conditioner's target temperature is 26℃, increasing by 0.1℃ with each adjustment; the heater is maintained at 26℃;

[0139] If the temperature stabilizes (approaching the target temperature value at an appropriate rate), the control result output by the large control model can be used to maintain the original control of the air conditioner and heater.

[0140] Taking temperature as an example, "″" indicates the beginning and end of the prompt word. The template for the first prompt word is as follows:

[0141] "″"

[0142] You are an indoor air analysis expert who generates a one-sentence temperature rating by combining indoor temperature data (currentTemperature), a list of devices in the room (deviceList), device location information (deviceLocation), historical data of devices in the room (deviceHistory), historical data curves of room air quality (currentTemperatureLine), last temperature evaluation (lastEvaluation), and indoor temperature target range (confortableRange) reported by current smart home appliances.

[0143] Example:

[0144] Input:{"device":"Thermometer","currentTemperature":"23","deviceList":["Thermometer","Wall-mounted Air Conditioner"],"deviceLocation":{"Thermometer":"Center of Room","Wall-mounted Air Conditioner":"By Window"},"deviceHistory":{"Thermometer":"22.6","Wall-mounted Air Conditioner":"21.3"},"currentTemperatureLine":["22.1","22.4"],"lastEvaluation":"The temperature is stabilizing; the temperature by the window is lower than the temperature in the center of the room.","confortableRange":{"max":"26","min":"Wall-mounted Air Conditioner"}}

[0145] Output: The temperature is stabilizing; the temperature near the window is lower than the temperature in the center of the room.

[0146] "″"

[0147] The template for the second prompt word is as follows:

[0148] "″"

[0149] You are a device control expert. Based on the current room temperature evaluation, the target smart home appliance list (deviceList), deviceTypeFunctions, downstream device locations (deviceLocation), user preferences, user's historical device control records (controlHistory), user's geographic location (location), and temperature changes (tempChange), generate device control commands. These commands must be output as JSON strings and meet the following requirements.

[0150] 1. Output the relationship between devices and command sets in key-value pair format, for example: {"wall-mounted air conditioner":{"onOff":"true"}

[0151] 2. Commands and command values ​​in the device command set must be output strictly according to the commands, value ranges, and types specified in the device function set (deviceTypeFunctions). For example, the command to turn on an air conditioner is {"onOff":"true"}.

[0152] 3. Downlink devices that do not participate in equipment control may not appear in the equipment control commands.

[0153] Example:

[0154] Input:{"location":"Qingdao","deviceList":["wall-mounted air conditioner","heater"],"deviceLocation":{"wall-mounted air conditioner":"by the window","heater":"inside the room near the door"},"deviceTypeFunctions":{"wall-mounted air conditioner":{"targetTemp":{"range":{"max":30,"min":16,"step":1},"type":"int"}}},"preference":["warm","long-term optimization","night control"],"evaluation":"The temperature is stabilizing; the temperature near the window is lower than the temperature in the center of the room","controlHistor y":[{"targetTemp":"26","device":"wall-mounted air conditioner"}],"tempChange":"0.1"}

[0155] Output:{"wall-mounted air conditioner":["targetTemp":"26","pressure":"5"],"heater":["targetTemp":"26"]}

[0156] "″"

[0157] In this context, example represents an example, Input represents input, Output represents output, device represents a device, onOff represents an on / off switch, true represents the on / off state of the on / off switch, range represents a range, max represents the maximum value (30℃ in the example of the second prompt word template), min represents the minimum value (16℃ in the example of the second prompt word template), step represents each temperature adjustment (1℃ in the example of the second prompt word template), targetTemp represents the target temperature, type represents the data type of the target temperature, int represents an integer variable, and pressure represents the compressor pressure.

[0158] To better meet specific user needs, the timing and duration of temperature adjustments can be set by the user (e.g., the user expects rapid warming at night and slow warming during the day). This allows the control model to be configured with control commands closer to user needs, enhancing the user experience.

[0159] In one embodiment, the large control model needs to be trained before it can obtain control results based on the second prompt word and the input:

[0160] For example, a training sample set is obtained, and the control model is trained based on the training sample set;

[0161] The training samples in the training sample set include analysis result samples, control command samples from the previous cycle, target parameter value samples, and control command samples from the current cycle.

[0162] When the analysis result sample represents the target environmental parameter approaching the target parameter value sample at a preset speed, the control command sample of the current cycle is the same as the control command sample of the previous cycle.

[0163] The large analysis model and the large control model are trained separately. During the training of both models, the LoRA model can be used to fine-tune the large model. During training, the large model is trained using a training sample set. Starting from the impact of smart home appliances on environmental parameters, with the goal of controlling smart home appliances to achieve a suitable environment, the data types of the training sample set are obtained. For example, the functional control of smart home appliances to adjust the environment, changes in environmental parameters, etc. More specific data can be found in the input of the large model in the above embodiments.

[0164] Controlling the output method of large models:

[0165] After obtaining the control parameters, the control model can output the control of smart home appliances in the form of key-value pairs to achieve the corresponding state; the commands and command values ​​in the control model need to be executed in the corresponding format.

[0166] Figure 5 A diagram of an intelligent control device based on a large model is provided for an embodiment of the present invention, such as... Figure 5 As shown, the device includes: an acquisition module 501, a first processing module 502, and a second processing module 503;

[0167] The acquisition module 501 is used to acquire the parameter values ​​of the target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter.

[0168] The first processing module 502 is used to input the parameter values, the indoor location information of at least some smart home appliances, and the first prompt word into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or change trend of the target environmental parameters based on the parameter values ​​and location information.

[0169] The second processing module 503 is used to input the analysis results and the second prompt word into the control model to obtain control results. The second prompt word is used to instruct the control model to generate control instructions for at least one target smart home appliance in the set of home appliances based on the analysis results. The control instructions are used to control the target smart home appliance to adjust the target environmental parameters.

[0170] This application also provides an electronic device, including: at least one processor and a memory;

[0171] The memory stores computer-executed instructions;

[0172] The at least one processor executes computer execution instructions stored in the memory, causing the at least one processor to execute an intelligent control method based on a large model.

[0173] Figure 6 This is a hardware schematic diagram of an electronic device provided in an embodiment of the present invention. For example... Figure 6 As shown, the electronic device 60 provided in this embodiment includes at least one processor 601 and a memory 602. The device 60 also includes a communication component 603. The processor 601, memory 602, and communication component 603 are connected via a bus 604.

[0174] In the specific implementation process, at least one processor 601 executes the computer execution instructions stored in the memory 602, causing at least one processor 601 to execute the above-mentioned intelligent control method based on a large model.

[0175] The specific implementation process of processor 601 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0176] In the above Figure 6 In the illustrated embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0177] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0178] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0179] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the method described above.

[0180] The aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0181] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0182] The division of units described herein is merely a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0183] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0184] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0185] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A smart control method based on a large model, characterized in that, For controlling a collection of home appliances, the collection including multiple smart home appliances, the smart home appliances being used to collect and / or adjust environmental parameters, the method includes: Obtain the parameter values ​​of the target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter; The parameter values, the indoor location information of at least some smart home appliances, and the first prompt word are input into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameters based on the parameter values ​​and location information. The analysis results and the second prompt word are input into the control model to obtain control results. The second prompt word is used to instruct the control model to generate control instructions for at least one target smart home appliance in the set of home appliances based on the analysis results. The control instructions are used to control the target smart home appliance to adjust the target environmental parameters.

2. The method according to claim 1, characterized in that, The parameter values, the indoor location information of at least some of the smart home appliances, and the first prompt word are input into the large-scale analysis model to obtain the analysis results, including: Every preset period, input the following information and the first prompt word into the large analysis model to obtain the analysis results for the current period: The current parameter value and historical parameter value corresponding to the current period; The location information of at least some of the smart home appliances indoors; Analysis results from the previous period; The target parameter values ​​corresponding to the target environmental parameters; Specifically, the first prompt word is used to indicate: based on the parameter values ​​and location information of the current period, combined with the target parameter values ​​and the analysis results of the previous period, analyze the indoor distribution and / or trend of the target environmental parameters in the current period.

3. The method according to claim 1, characterized in that, The analysis results and the second prompt word are input into the large-scale control model to obtain the control results, including: Every preset period, the following information and the second prompt word are input into the control model to obtain the control result for the current period: Analysis results corresponding to the current cycle; The control command corresponding to the previous cycle; The target parameter values ​​corresponding to the target environmental parameters; The second prompt word is specifically used to indicate: based on the analysis results of the current cycle, combined with the target parameter value and the control command of the previous cycle, the control result of the current cycle is obtained; The target parameter value can be set in any of the following ways: default setting for smart home appliances, or user setting for smart home appliances.

4. The method according to claim 3, characterized in that, The information input into the control model also includes: The list of smart home appliances in the home appliance collection; Each smart home appliance has its own set of functions; each function in the set is used to adjust one or more environmental parameters. The second prompt word is also used to indicate: based on the function sets corresponding to each smart home appliance, at least one target smart home appliance is determined from the set of appliances to adjust the target environmental parameters.

5. The method according to claim 3, characterized in that, The method further includes: Obtain a training sample set and train the large control model based on the training sample set; The training samples in the training sample set include analysis result samples, control command samples from the previous cycle, target parameter value samples, and control command samples from the current cycle. When the analysis result sample represents the target environmental parameter approaching the target parameter value sample at a preset speed, the control command sample of the current cycle is the same as the control command sample of the previous cycle.

6. The method according to any one of claims 3-5, characterized in that, The parameter values ​​are indoor parameter values; the information input to the control model also includes at least one of the following: The target environmental parameter is the outdoor parameter value, wherein the outdoor parameter value is obtained by querying the geographical location information of the appliance set. User preference information includes target parameter values, the speed at which the target environmental parameters approach the target parameter values, and the usage time periods of the respective functional groups of each smart home appliance.

7. An intelligent control device based on a large model, characterized in that, For controlling a collection of home appliances, the collection including multiple smart home appliances, the smart home appliances being used to collect and / or adjust environmental parameters, the device comprising: The acquisition module is used to acquire the parameter values ​​of target environmental parameters collected by at least some of the smart home appliances in the home appliance set; the target environmental parameter is any environmental parameter that can be collected by the home appliance set, and the parameter value includes the historical parameter value and / or the current parameter value of the target environmental parameter; The first processing module is used to input the parameter values, the location information of at least some smart home appliances in the room, and the first prompt word into the analysis model to obtain the analysis results. The first prompt word is used to instruct the analysis model to analyze the indoor distribution and / or changing trend of the target environmental parameters based on the parameter values ​​and location information. The second processing module is used to input the analysis results and the second prompt word into the control model to obtain the control result. The second prompt word is used to instruct the control model to generate a control command for at least one target smart home appliance in the set of home appliances based on the analysis results. The control command is used to control the target smart home appliance to adjust the target environmental parameters.

8. An electronic device, characterized in that, include: At least one processor and memory; The memory stores computer-executed instructions; The at least one processor executes computer execution instructions stored in the memory, causing the at least one processor to perform the intelligent control method based on a large model as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent control method based on a large model as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes instructions that, when executed on an electronic device, cause the electronic device to implement the intelligent control method based on a large model as described in any one of claims 1-6.

Images