Intelligent device control method and system

Abstract

The present invention discloses a smart device control method and system, and relates to the field of smart control technology. The smart device control system includes: a usage scenario model acquisition module, a control scenario construction module, a control scenario optimization module, a real-time control module, an anomaly detection module and a visual interface module. The present invention analyzes the control record samples of multiple smart devices, mines the correlation between devices, and implements a linkage control strategy for multiple devices. This strategy enables different devices to work together in specific scenarios or conditions, providing a more complex and flexible control experience; by analyzing the time patterns and habitual control strategies of associated record samples, the user's usage habits at different times and scenarios are identified, thereby intelligently optimizing the control strategy, enabling the system to adapt to the user's behavior pattern, and providing the user with a more convenient and personalized control solution.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent device control method and system. Background Art

[0002] With the rapid development of the Internet of Things and smart device technologies, more and more smart devices are being used in home, office, and industrial scenarios. These smart devices can usually be interconnected through wireless networks, supporting remote control, automated operations, and linkage with other devices. However, the existing smart device control systems still have many shortcomings in terms of optimization of control logic, linkage between devices, and convenience of operation. For example, most systems only support simple timing control or trigger control based on specific conditions, and lack support for complex control strategies between multiple devices. In addition, when devices of different brands and models are interconnected, it is difficult to achieve unified management and linkage control due to different communication protocols and control standards.

[0003] Current smart device control methods are usually based on preset control instructions and single-scenario control, which makes it difficult to meet users' refined control needs in changing scenarios. The diverse working scenarios of smart devices require the system to have real-time data processing capabilities and be able to dynamically adjust control strategies based on environmental changes and user usage habits. Especially in home and office environments, the collaborative use of different devices is gradually increasing. How to achieve intelligent linkage of multiple devices and improve user experience has become a hot issue in the research of intelligent control technology.

[0004] Therefore, there is an urgent need for an intelligent device control method and system that can integrate multi-device information, automatically optimize control strategies, and perform real-time monitoring and exception handling to meet users' diverse needs for smart scenes and improve the control efficiency and safety of smart devices. Summary of the invention

[0005] A method for controlling an intelligent device comprises the following steps:

[0006] Step S1: The user inputs the size information of the workspace, matches it in the workspace template library, and obtains the usage scenario model of the smart device. The usage scenario model is composed of the complete workspace of the user using the smart device. The usage scenario model includes the function identification of each area in the workspace, environmental information and smart device information;

[0007] Step S2: The user adds multiple smart devices to the usage scenario model through the control terminal, and improves the smart device information of the usage scenario model. The smart device establishes a connection with the control system through a wireless network. The control system obtains parameter information of the smart device through the model information of the smart device, and uniquely marks each smart device.

[0008] Step S3: define multiple control scenarios on the control terminal, each scenario includes a group of smart devices and corresponding control instructions, and establish a unique control tag for each control instruction of each smart device; the user obtains multiple required control scenarios by adjusting the preset control scenario template; obtain an associated record sample set, each associated record sample corresponds to a control instruction generated by a smart device, the associated record sample is composed of a control tag and a record time point, the associated record sample set includes associated record samples generated by all smart devices over time, and the associated record sample set is sent to the control scenario optimization model for calculation to obtain a linkage control strategy and a habitual control strategy, and the linkage control strategy and the habitual control strategy are used as the control scenario optimization strategy and displayed on the control terminal, and the user updates the control scenario by selecting and determining the control scenario optimization strategy; the control scenario optimization model uses a sliding window to perform statistics on the associated record samples, performs record filtering and association filtering on the statistical results to obtain statistical records, and obtains the linkage control strategy according to the sorting results of the statistical records; calculates the regularity value of the associated record samples, and obtains the habitual control strategy according to the regularity value sorting;

[0009] Step S4: acquiring the status information of each smart device, the environment information of the workspace and the weather information of the location in real time, and displaying the current status information of the smart device and the environment information of the workspace through the control terminal; using the status information of each smart device, the environment information of the workspace and the weather information of the location as adjustment control sample data, and sending them into the real-time control model for calculation to obtain adjustment instructions; the real-time control model is trained by historical data to obtain the correlation between the control adjustment sample data and the adjustment instructions; calculating the fitness R of the control instruction and the adjustment instruction, comparing the fitness of the control instruction and the adjustment instruction, and selecting the instruction with the largest fitness to control the smart device;

[0010] Step S5: Perform abnormality detection on each smart device according to the status information of the smart device and the environmental information of the workspace to obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the smart device, and send an alarm notification to the control terminal; when performing emergency control, automatically save the working status of the smart device before the emergency control, and the saved record is used to restore the working status of the smart device; the user remotely views and controls the status of the smart device on the control terminal through manual control or voice interaction.

[0011] As a preferred technical solution of the present invention, obtaining the usage scenario model includes:

[0012] Construction of workspace template library: Get the type of workspace, and build multiple preset workspace templates based on the type. The workspace template is the construction parameter of the use scenario model, which is displayed through a two-dimensional spatial top view. In the template, the workspace is divided into multiple areas based on the physical partition of the space and the recognition of artificial areas. A functional identifier is added to each area. The functional identifier is not unique, and the same area has multiple functional identifiers. The preset workspace template is added to the workspace template library, and the workspace templates are classified by type in the workspace template library;

[0013] Matching workspace template: The user selects the required workspace category on the control terminal and enters the size information of the workspace. The size information is used to match the workspace template with the closest size in the workspace template library.

[0014] Parameter adjustment of the workspace template: adjust the matched workspace template according to the input size information to obtain the workspace template of the size required by the user;

[0015] Build a usage scenario model: Automatically build a usage scenario model based on the final workspace template. The usage scenario model is a three-dimensional space model that includes segmented areas. It is displayed on the control terminal through a 3D model and a top view of the model.

[0016] As a preferred technical solution of the present invention, improving the usage scenario model includes:

[0017] The user adds a smart device on the control terminal by selecting the smart device category or smart device model under the category in the smart device list, and dragging it to the location in the usage scenario model to add it. If the category of the smart device is selected for addition, the model needs to be selected after the addition is completed; or directly click the location in the usage scenario model to add a smart device, and enter the identification information of the smart device to match the model;

[0018] After completing the addition of smart devices, obtain the parameter information of the smart devices according to their models, calculate the spatial parameters of the corresponding area of ​​each smart device in the usage scenario model, compare the spatial parameters of the area with the parameter information of the smart devices, and perform the comparison through the standard device parameter standard spatial parameter table. If the comparison result deviation exceeds the set threshold, a prompt will be given on the control terminal;

[0019] The complete usage scenario model includes the function identification, environmental information and smart device information of each area in the workspace. The environmental information is obtained through weather information and sensors. In the usage scenario model, the function identification, environmental information and smart device information are displayed by selecting environmental parameters or through the area list.

[0020] As a preferred technical solution of the present invention, the control scenario includes:

[0021] There are three types of control scenarios: specific trigger scenarios, time status scenarios, and custom scenarios. Specific trigger scenarios are triggered by specific events; time status scenarios are triggered by time, specifically distinguished by month, day, and hour; custom scenarios are customized by users; the three types of control scenarios are compatible. If there is a situation of controlling the same smart device, the adaptability of the control command is calculated, and the command with higher adaptability is selected to control the smart device;

[0022] The user selects a preset control scene template through the control terminal. The template includes a set of preset smart devices and corresponding control instructions. The user chooses to directly use the preset template or make adjustments based on the preset template.

[0023] As a preferred technical solution of the present invention, the control scenario optimization model includes:

[0024] The data recording layer is used to record the control instructions of each smart device and convert the records into the format of associated record samples. The associated record samples are composed of control tags and control time. The associated record samples are connected in chronological order to obtain an associated record sample set.

[0025] The association statistics layer is used to obtain statistical records of association record sample combinations that meet the requirements in the association record sample set, and obtains them by sliding window statistics; it includes the following steps:

[0026] Step A1: Perform the first round of iteration, set the sliding window to 2 bits, use the sliding window to count the adjacent association record sample combinations in the association record sample one by one, obtain the statistical results, perform record filtering and association filtering on the statistical results, and record the filtered statistical results into statistical records, wherein the statistical records are sets of sets, each set is the statistical result of one round of iteration, and the statistical results are a data string of combination-number-data position of each association record sample combination;

[0027] Step A101: Filter the associated record sample combination of the first iteration; calculate the time difference between the associated record samples corresponding to the first and second positions in the sliding window. If the time difference exceeds the dynamic association threshold C, do not count the associated record sample combination in the current sliding window;

[0028] The calculation formula of dynamic correlation threshold C is: , where C0 represents the initial threshold, n represents the number of iterations of the sliding window, the number of iterations n in the first round of iterations is recorded as 0, and the value of n is increased by 1 in each round of iterations, C M Indicates the maximum threshold adjustment value;

[0029] Step A102: Send the statistical results after record filtering to the correlation calculation layer to calculate the correlation, and perform correlation filtering on the statistical results, and retain the correlation record sample combinations whose correlation meets the standard in the statistical results, where the correlation of the correlation record sample combination is greater than a set threshold;

[0030] Step A103: Record the set corresponding to the statistical result after the association filtering into the statistical record, completing the first round of sliding window iteration;

[0031] Step A2: perform the next iteration, use the associated record sample combination recorded in the last sliding window as the first position of the sliding window, find the position of the associated record sample combination through the data position in the last round of statistical results, and count the adjacent n+2 associated record sample combinations in the associated record samples one by one, perform record filtering and association filtering on the statistical results, record the set corresponding to the filtered statistical results into the statistical record, and complete this round of iteration;

[0032] Step A3: repeat the iterative process in step A2 until there is no associated record sample combination with a satisfactory association degree in the iteration;

[0033] The correlation calculation layer is used to calculate the correlation L of the associated record sample combination in the statistical results, the correlation L of the i-th associated record sample combination i The calculation formula is: , where N i is the number of times the i-th associated record sample combination appears in the statistical record, T represents the number of days, M is the number of all associated record sample combinations in this iteration, represents the proportion of the i-th associated record combination, j represents the j-th record of the i-th associated record combination, t represents the control time corresponding to the j-th record, d() is the Euclidean distance calculation function, and d(j,t) represents the time similarity of the i-th associated record combination; A and B are the adjustment weights of the proportion and time similarity of the associated record combination, respectively. The initial weight is set to A=B=1 / 2, and A+B=1 is satisfied;

[0034] The regularity value calculation layer is used to calculate the regularity value of the associated record samples. It is obtained by counting the number of associated record samples in units of days and months. The calculation formula is: , where P d and P m Represent the regularity value of the day and the regularity value of the month, P d By counting the number of associated record samples occurring in the same time period in units of days, i The ratio of the total number of days T is obtained, P m By counting the number of related record samples occurring on the same day in units of months, m di and the total number of months T mThe maximum value of P is 1;

[0035] The strategy summary layer is used to obtain the associated record sample combinations and associated record samples with regular values ​​in the statistical records, and sort them respectively. The associated record sample combinations that are at the top of the sort are taken as the linkage control strategy output, and the associated record samples that are at the top of the sort are taken as the habitual control strategy output.

[0036] As a preferred technical solution of the present invention, the real-time control model includes:

[0037] By acquiring historical data, which includes the status information of smart devices, the environment information of the workspace, and the weather information of the location, and recording the control instructions as classification labels in the historical data, a real-time control data set is constructed through the historical data, and the data set is divided into a training set, a validation set, and a test set;

[0038] The real-time control training set and verification set are sent to the classification model based on pre-training, and the control instructions are used as the classification target for training to obtain the initial real-time control model;

[0039] The real-time control test set is sent to the initial real-time control model for accuracy evaluation. If the accuracy is within the set range, the model is output as a real-time control model. Otherwise, the hyperparameters of the initial real-time control model are adjusted and the model is retrained until the accuracy meets the standard.

[0040] As a preferred technical solution of the present invention, the fitness calculation includes:

[0041] When there are control instructions and adjustment instructions to control the same device, the adaptability R between the instruction and the device is calculated. The adaptability R between the instruction and the device is calculated by obtaining the standard state of the environmental information of the current device in its working area. The adaptability R is the inverse of the distance difference between the standard state, the current state and the state corresponding to the instruction in the device state vector matrix. The smaller the distance difference, the higher the adaptability, and the corresponding instruction is closer to the standard state; the standard state is searched in the device environment table.

[0042] As a preferred technical solution of the present invention, anomaly detection includes:

[0043] Anomaly detection uses a multi-dimensional anomaly detection model, which comprehensively considers the status information of smart devices, the environmental data of the workspace, and the user's historical operation records for anomaly detection. The anomaly detection results include three categories: device operation anomalies, environmental anomalies, and multi-device linkage anomalies. According to different anomaly types, corresponding emergency instructions are generated, and the working status of smart devices is adjusted in real time.

[0044] An intelligent device control system includes the following modules:

[0045] The usage scenario model acquisition module includes a preset template construction unit, a template matching unit, a model generation unit and a model improvement unit. The preset template construction unit is used to build and add preset workspace templates, including background addition and foreground addition. The templates added in the background are displayed to all users by the administrator, and the templates added in the foreground are performed by the user and are only displayed to the user. The user recommends the template added in the foreground, and the template is added to the background after being reviewed and approved by the administrator; the template matching unit is used to match in the workspace template library, and the size information and category of the workspace need to be input; the model generation unit is used to automatically adjust the parameters of the matched workspace template according to the parameters input by the user, and generate a three-dimensional usage scenario model. This unit manually adjusts the workspace template by the user; the model improvement unit is used to add smart devices and mark the regional functions to the usage scenario model, and obtain a usage scenario model including the functional identification of each area in the workspace, environmental information and smart device information;

[0046] The control scenario construction module includes a scenario construction unit and a scenario management unit. The scenario construction unit is used to obtain multiple required control scenarios by adjusting the preset control scenario template; the scenario management unit is used to adjust the control scenario and select the control scenario optimization strategy output by the control scenario optimization module for automatic adjustment;

[0047] The control scenario optimization module is composed of a control scenario optimization model, and is used to send the associated record sample set into the control scenario optimization model for calculation, perform statistics on the associated record samples using sliding window statistics, perform record filtering and associated filtering on the statistical results to obtain statistical records, and obtain the linkage control strategy according to the sorting results of the statistical records; calculate the regularity value of the associated record samples, and obtain the habitual control strategy according to the regularity value sorting; use the linkage control strategy and the habitual control strategy as the control scenario optimization strategy, and display them on the control terminal;

[0048] The real-time control module includes an information acquisition unit, an adjustment control unit and a control verification unit. The information acquisition unit is used to acquire the status information of each smart device, the environmental information of the workspace and the weather information of the location in real time. The adjustment control unit is composed of a real-time control model, which is used to use the status information of each smart device, the environmental information of the workspace and the weather information of the location as adjustment control sample data, and send them into the real-time control model for calculation to obtain adjustment instructions; the control verification unit is used to calculate the fitness of the control instruction and the adjustment instruction, compare the fitness of the control instruction and the adjustment instruction, and select the instruction with the large fitness to control the smart device;

[0049] The abnormality detection module includes an emergency control unit and a record recovery unit. The emergency control unit is used to perform abnormality detection on each intelligent device according to the status information of the intelligent device and the environmental information of the workspace, obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the intelligent device, and send an alarm notification to the control terminal; the record recovery unit is used to automatically save the working state of the intelligent device before the emergency control when performing emergency control, and the saved record is used to restore the working state of the intelligent device;

[0050] The visualization interface module is used to visualize the system, and the displayed contents include the main interface, model acquisition interface, usage scenario interface and instruction adjustment interface; the main interface is used to provide the user with overview information of the control workspace; the model acquisition interface is used to provide the user with an operation interface for model acquisition; the usage scenario interface is used to provide the user with an interface for usage scenario adjustment; the instruction adjustment interface is used to provide the user with an interface for adjusting instructions.

[0051] The present invention has the following advantages:

[0052] The present invention analyzes the control record samples of multiple smart devices, explores the correlation between devices, and implements a linkage control strategy for multiple devices. This strategy enables different devices to work together in specific scenarios or conditions, providing a more complex and flexible control experience; by analyzing the time pattern and habit control strategy of the associated record samples, the user's usage habits at different times and scenarios are identified, thereby intelligently optimizing the control strategy, enabling the system to adapt to the user's behavior pattern, and providing users with a more convenient and personalized control solution.

[0053] The present invention monitors the environmental information, equipment status and weather information of the workspace in real time, and uses the real-time control model trained with historical data to make dynamic adjustments, thereby ensuring that the smart device can operate in different environments, automatically adjusting the device parameters according to the real-time environment, and improving the system adaptability and stability.

[0054] The present invention uses a multi-dimensional anomaly detection model to conduct in-depth analysis of the device operating status, environmental conditions, and user operations, quickly identify abnormal situations, and generate emergency control instructions. This mechanism not only improves the security of the system, but also provides more comprehensive protection for intelligent control; through the management and optimization of multiple control scenarios, it supports users to switch between different scenarios or create custom scenarios. Whether it is triggered by specific events, time triggers, or user-defined settings, it can flexibly adapt to the needs of various scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a schematic diagram of the structure of an intelligent device control system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0057] Embodiment 1, a method for controlling an intelligent device, comprising the following steps:

[0058] Step S1: The user inputs the size information (length, width, height) of the workspace, matches it in the workspace template library, and obtains the usage scenario model of the smart device. The usage scenario model consists of the complete workspace of the user using the smart device, and the usage scenario model includes the function identification of each area in the workspace, environmental information and smart device information;

[0059] Obtaining the usage scenario model includes:

[0060] Construction of workspace template library: Get the type of workspace, and build multiple preset workspace templates based on the type. The workspace template is the construction parameter of the use scenario model, which is displayed through a two-dimensional spatial top view. In the template, the workspace is divided into multiple areas based on the physical partition of the space and the recognition of artificial areas. A functional identifier is added to each area. The functional identifier is not unique, and the same area has multiple functional identifiers. The preset workspace template is added to the workspace template library, and the workspace templates are classified by type in the workspace template library;

[0061] Matching workspace template: The user selects the required workspace category on the control terminal and enters the size information of the workspace. The size information is used to match the workspace template with the closest size in the workspace template library.

[0062] Parameter adjustment of the workspace template: adjust the matched workspace template according to the input size information to obtain the workspace template of the size required by the user;

[0063] Build a usage scenario model: Automatically build a usage scenario model based on the final workspace template. The usage scenario model is a three-dimensional space model that includes segmented areas. It is displayed on the control terminal through a 3D model and a top view of the model.

[0064] Step S2: The user adds multiple smart devices to the usage scenario model through the control terminal, and improves the smart device information of the usage scenario model. The smart device establishes a connection with the control system through a wireless network. The control system obtains parameter information of the smart device through the model information of the smart device, and uniquely marks each smart device.

[0065] Improve the usage scenario model including:

[0066] The user adds a smart device on the control terminal by selecting the smart device category or smart device model under the category in the smart device list, and dragging it to the location in the usage scenario model to add it. If the category of the smart device is selected for addition, the model needs to be selected after the addition is completed; or directly click the location in the usage scenario model to add a smart device, and enter the identification information of the smart device to match the model;

[0067] After completing the addition of smart devices, obtain the parameter information of the smart devices according to their models, calculate the spatial parameters of the corresponding area of ​​each smart device in the usage scenario model, compare the spatial parameters of the area with the parameter information of the smart devices, and perform the comparison through the standard device parameter standard spatial parameter table. If the comparison result deviation exceeds the set threshold, a prompt will be given on the control terminal;

[0068] The complete usage scenario model includes the function identification, environmental information and smart device information of each area in the workspace. The environmental information is obtained through weather information and sensors. In the usage scenario model, the function identification, environmental information and smart device information are displayed by selecting environmental parameters or through the area list.

[0069] Step S3: define multiple control scenarios on the control terminal, each scenario includes a group of smart devices and corresponding control instructions, and establish a unique control tag for each control instruction of each smart device; the user obtains multiple required control scenarios by adjusting the preset control scenario template; obtain an associated record sample set, each associated record sample corresponds to a control instruction generated by a smart device, the associated record sample is composed of a control tag and a record time point, the associated record sample set includes associated record samples generated by all smart devices over time, and the associated record sample set is sent to the control scenario optimization model for calculation to obtain a linkage control strategy and a habitual control strategy, and the linkage control strategy and the habitual control strategy are used as the control scenario optimization strategy and displayed on the control terminal, and the user updates the control scenario by selecting and determining the control scenario optimization strategy; the control scenario optimization model uses a sliding window to perform statistics on the associated record samples, performs record filtering and association filtering on the statistical results to obtain statistical records, and obtains the linkage control strategy according to the sorting results of the statistical records; calculates the regularity value of the associated record samples, and obtains the habitual control strategy according to the regularity value sorting;

[0070] Control scenarios include:

[0071] The data recording layer is used to record the control instructions of each smart device and convert the records into the format of associated record samples. The associated record samples are composed of control tags and control time. The associated record samples are connected in chronological order to obtain an associated record sample set.

[0072] The association statistics layer is used to obtain statistical records of association record sample combinations that meet the requirements in the association record sample set, and obtains them by sliding window statistics; it includes the following steps:

[0073] Step A1: Perform the first round of iteration, set the sliding window to 2 bits, use the sliding window to count the adjacent association record sample combinations in the association record sample one by one, obtain the statistical results, perform record filtering and association filtering on the statistical results, and record the filtered statistical results into statistical records, wherein the statistical records are sets of sets, each set is the statistical result of one round of iteration, and the statistical results are a data string of combination-number-data position of each association record sample combination;

[0074] Step A101: Filter the associated record sample combination of the first iteration; calculate the time difference between the associated record samples corresponding to the first and second positions in the sliding window. If the time difference exceeds the dynamic association threshold C, do not count the associated record sample combination in the current sliding window;

[0075] The calculation formula of dynamic correlation threshold C is: , where C0 represents the initial threshold, n represents the number of iterations of the sliding window, the number of iterations n in the first round of iterations is recorded as 0, and the value of n is increased by 1 in each round of iterations, C M Indicates the maximum threshold adjustment value;

[0076] Step A102: Send the statistical results after record filtering to the correlation calculation layer to calculate the correlation, and perform correlation filtering on the statistical results, and retain the correlation record sample combinations whose correlation meets the standard in the statistical results, where the correlation of the correlation record sample combination is greater than a set threshold;

[0077] Step A103: Record the set corresponding to the statistical result after the association filtering into the statistical record, completing the first round of sliding window iteration;

[0078] Step A2: perform the next iteration, use the associated record sample combination recorded in the last sliding window as the first position of the sliding window, find the position of the associated record sample combination through the data position in the last round of statistical results, and count the adjacent n+2 associated record sample combinations in the associated record samples one by one, perform record filtering and association filtering on the statistical results, record the set corresponding to the filtered statistical results into the statistical record, and complete this round of iteration;

[0079] Step A3: repeat the iterative process in step A2 until there is no associated record sample combination with a satisfactory association degree in the iteration;

[0080] The correlation calculation layer is used to calculate the correlation L of the associated record sample combination in the statistical results, the correlation L of the i-th associated record sample combination i The calculation formula is: , where N i is the number of times the i-th associated record sample combination appears in the statistical record, T represents the number of days, M is the number of all associated record sample combinations in this iteration, represents the proportion of the i-th associated record combination, j represents the j-th record of the i-th associated record combination, t represents the control time corresponding to the j-th record, d() is the Euclidean distance calculation function, and d(j,t) represents the time similarity of the i-th associated record combination; A and B are the adjustment weights of the proportion and time similarity of the associated record combination, respectively. The initial weight is set to A=B=1 / 2, and A+B=1 is satisfied;

[0081] The regularity value calculation layer is used to calculate the regularity value of the associated record samples. It is obtained by counting the number of associated record samples in units of days and months. The calculation formula is: , where P d and P m Represent the regularity value of the day and the regularity value of the month, P d By counting the number of associated record samples occurring in the same time period in units of days, i The ratio of the total number of days T is obtained, P m By counting the number of related record samples occurring on the same day in units of months, m di and the total number of months T m The maximum value of P is 1;

[0082] The strategy summary layer is used to obtain the associated record sample combinations and associated record samples with regular values ​​in the statistical records, and sort them respectively. The associated record sample combinations that are at the top of the sort are taken as the linkage control strategy output, and the associated record samples that are at the top of the sort are taken as the habitual control strategy output.

[0083] The sorting method of the regularity value is to sort directly according to the numerical value. The sorting method of the associated record sample combination is to sort the evaluation value K of the associated sample combination numerically. The calculation formula of the evaluation value K is: , where L is the correlation degree, NOR() is the normalization function, n is the number of associated record sample combinations, o is the number of associated record samples in the associated record sample combination, w1 and w2 are the weighted values ​​of n and o respectively;

[0084] Step S4: acquiring the status information of each smart device, the environment information of the workspace and the weather information of the location in real time, and displaying the current status information of the smart device and the environment information of the workspace through the control terminal; using the status information of each smart device, the environment information of the workspace and the weather information of the location as adjustment control sample data, and sending them into the real-time control model for calculation to obtain adjustment instructions; the real-time control model is trained by historical data to obtain the correlation between the control adjustment sample data and the adjustment instructions; calculating the fitness R of the control instruction and the adjustment instruction, comparing the fitness of the control instruction and the adjustment instruction, and selecting the instruction with the largest fitness to control the smart device;

[0085] The real-time control model includes:

[0086] By acquiring historical data, which includes the status information of smart devices, the environment information of the workspace, and the weather information of the location, and recording the control instructions as classification labels in the historical data, a real-time control data set is constructed through the historical data, and the data set is divided into a training set, a validation set, and a test set;

[0087] The real-time control training set and verification set are sent to the classification model based on pre-training, and the control instructions are used as the classification target for training to obtain the initial real-time control model;

[0088] The pre-trained classification model has been trained with a standard data set and has good classification results. The required initial real-time control model is constructed through the pre-trained classification model using transfer learning.

[0089] The real-time control test set is sent to the initial real-time control model for accuracy evaluation. If the accuracy is within the set range, the model is output as a real-time control model. Otherwise, the hyperparameters of the initial real-time control model are adjusted and the model is retrained until the accuracy meets the standard.

[0090] The fitness calculation includes:

[0091] When there are control instructions and adjustment instructions to control the same device, the adaptability R between the instruction and the device is calculated. The adaptability R between the instruction and the device is calculated by obtaining the standard state of the environmental information of the current device in its working area. The adaptability R is the inverse of the distance difference between the standard state, the current state and the state corresponding to the instruction in the device state vector matrix. The smaller the distance difference, the higher the adaptability, and the corresponding instruction is closer to the standard state; the standard state is searched in the device environment table.

[0092] Step S5: Perform abnormality detection on each smart device according to the status information of the smart device and the environmental information of the workspace to obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the smart device, and send an alarm notification to the control terminal; when performing emergency control, automatically save the working status of the smart device before the emergency control, and the saved record is used to restore the working status of the smart device; the user remotely views and controls the status of the smart device on the control terminal through manual control or voice interaction.

[0093] Anomaly detection includes:

[0094] The anomaly detection module adopts a multi-dimensional anomaly detection model, which comprehensively considers the status information of smart devices, the environmental data of the workspace, and the user's historical operation records for detection. The anomaly detection results include three categories: device operation anomalies, environmental anomalies, and multi-device linkage anomalies. According to different anomaly types, corresponding emergency instructions are generated, and the working status of smart devices is adjusted in real time to ensure the safety of equipment and working environment.

[0095] Example 2, a smart device control system, see Figure 1 As shown, it includes the following modules:

[0096] The usage scenario model acquisition module includes a preset template construction unit, a template matching unit, a model generation unit and a model improvement unit. The preset template construction unit is used to build and add preset workspace templates, including background addition and foreground addition. The templates added in the background are displayed to all users by the administrator, and the templates added in the foreground are performed by the user and are only displayed to the user. The user recommends the template added in the foreground, and the template is added to the background after being reviewed and approved by the administrator; the template matching unit is used to match in the workspace template library, and the size information and category of the workspace need to be input; the model generation unit is used to automatically adjust the parameters of the matched workspace template according to the parameters input by the user, and generate a three-dimensional usage scenario model. This unit manually adjusts the workspace template by the user; the model improvement unit is used to add smart devices and mark the regional functions to the usage scenario model, and obtain a usage scenario model including the functional identification of each area in the workspace, environmental information and smart device information;

[0097] The control scenario construction module includes a scenario construction unit and a scenario management unit. The scenario construction unit is used to obtain multiple required control scenarios by adjusting the preset control scenario template; the scenario management unit is used to adjust the control scenario and select the control scenario optimization strategy output by the control scenario optimization module for automatic adjustment;

[0098] The control scenario optimization module is composed of a control scenario optimization model, and is used to send the associated record sample set into the control scenario optimization model for calculation, perform statistics on the associated record samples using sliding window statistics, perform record filtering and associated filtering on the statistical results to obtain statistical records, and obtain the linkage control strategy according to the sorting results of the statistical records; calculate the regularity value of the associated record samples, and obtain the habitual control strategy according to the regularity value sorting; use the linkage control strategy and the habitual control strategy as the control scenario optimization strategy, and display them on the control terminal;

[0099] The real-time control module includes an information acquisition unit, an adjustment control unit and a control verification unit. The information acquisition unit is used to acquire the status information of each smart device, the environmental information of the workspace and the weather information of the location in real time. The adjustment control unit is composed of a real-time control model, which is used to use the status information of each smart device, the environmental information of the workspace and the weather information of the location as adjustment control sample data, and send them into the real-time control model for calculation to obtain adjustment instructions; the control verification unit is used to calculate the fitness of the control instruction and the adjustment instruction, compare the fitness of the control instruction and the adjustment instruction, and select the instruction with the large fitness to control the smart device;

[0100] The abnormality detection module includes an emergency control unit and a record recovery unit. The emergency control unit is used to perform abnormality detection on each intelligent device according to the status information of the intelligent device and the environmental information of the workspace, obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the intelligent device, and send an alarm notification to the control terminal; the record recovery unit is used to automatically save the working state of the intelligent device before the emergency control when performing emergency control, and the saved record is used to restore the working state of the intelligent device;

[0101] The visualization interface module is used to visualize the system, and the displayed contents include the main interface, model acquisition interface, usage scenario interface and instruction adjustment interface; the main interface is used to provide the user with overview information of the control workspace; the model acquisition interface is used to provide the user with an operation interface for model acquisition; the usage scenario interface is used to provide the user with an interface for usage scenario adjustment; the instruction adjustment interface is used to provide the user with an interface for adjusting instructions.

[0102] Example 3, application of home intelligent control system:

[0103] In a home environment, users hope that smart lights, smart air conditioners, curtains, air purifiers and other devices can be automatically linked and controlled according to environmental information and user habits. The following are the specific steps:

[0104] Construction of usage scenario model: Users input the length, width, height and other dimensions of the family room through the control terminal, and the system automatically matches the usage scenario model of the family room from the workspace template library and adjusts its necessary parameters. The usage scenario model is displayed in 3D on the control terminal, including the functional identification of each area in the room, such as "resting area" and "entertainment area".

[0105] Adding smart devices: Users select and add smart devices in the home on the control terminal, such as smart lights, air conditioners, curtains, etc. The system obtains its parameters based on the device model information and automatically generates a unique identifier for the device to facilitate subsequent management and control.

[0106] Control scene definition and optimization: In the control terminal, users can preset different control scenes, such as "home mode" and "sleep mode". Each scene contains specific smart devices and control instructions. The system will analyze the linkage control strategy and habitual control strategy based on the user's previous operation records and device association samples. For example, in "sleep mode", the smart lights gradually dim, the air conditioner is controlled to a comfortable temperature, and the curtains are automatically drawn.

[0107] Real-time control and dynamic adjustment: The system collects the status information of smart devices, room temperature and humidity data, and weather conditions in real time, and uses real-time control models to analyze whether the device status needs to be adjusted. For example, when the temperature drops at night, the air conditioner automatically adjusts to low power consumption mode to ensure a comfortable ambient temperature.

[0108] Abnormal detection and emergency control: The system is equipped with an abnormal detection function. When an abnormality is detected in the equipment or environment, an emergency command is immediately triggered. For example, if the air quality is detected to be deteriorating, the system will automatically turn on the air purifier and send an alarm notification to the user.

[0109] Example 4, application of intelligent control system in industrial production workshop:

[0110] In an industrial production environment, precise control of factors such as temperature, humidity, and light is crucial to product quality. This embodiment demonstrates the application of an intelligent control system in a workshop.

[0111] Use scenario model generation: The administrator inputs the size of the production workshop and the workspace information, and the system finds a suitable template in the workspace template library and generates a 3D model of the workshop. The model includes production, storage and other areas, each with different functional identifiers.

[0112] Configuration of smart devices: Administrators select and configure smart temperature and humidity control systems, lighting systems, ventilation equipment, etc. The system obtains device parameter information through model information and marks the unique identity of each device to facilitate subsequent linkage management.

[0113] Construction and optimization of multiple control scenarios: Different production processes require different environmental conditions. The system sets scenarios such as "production mode" and "storage mode" for the workshop. By analyzing the associated record samples, the system can control multiple devices in a linked manner to achieve coordinated adjustment of temperature, humidity and lighting. For example, after entering the "storage mode", the temperature and humidity control system automatically adjusts to the storage environment standard, and the lighting system switches to energy-saving mode.

[0114] Real-time adjustment and dynamic optimization: The system obtains the environmental information of the workshop and the status data of each device in real time, and adjusts the environment according to the production progress. For example, if the temperature deviates from the standard value, the system automatically adjusts the operating speed of the ventilation equipment to ensure product quality.

[0115] Fault detection and emergency response: When a device fault or abnormal environmental parameter is detected, the system triggers an emergency response mechanism. For example, if the temperature and humidity exceed the warning range, the system will immediately shut down unnecessary equipment to reduce losses, and send abnormal information to the administrator for follow-up processing.

[0116] Example 5, application of office intelligent control system:

[0117] In the office, the intelligent system can realize intelligent control of air conditioning, lighting, curtains and other equipment to save energy and optimize the office environment:

[0118] Selection of workspace template: The user inputs the area and structure information of the office, the system selects the appropriate template from the space template library, and generates a 3D usage scenario model, including functional areas such as "meeting area", "work area", and "rest area".

[0119] Device management and addition: Users select and add smart devices in the system, such as air conditioners in office areas, lighting equipment in conference rooms, etc. The model information of each device is automatically identified by the system, and the parameters are recorded in the usage scenario model.

[0120] Optimize control scenarios: Set multiple scenarios according to the needs of the office space, including "working time mode", "lunch break mode" and "conference mode". Through the control scenario optimization model, the system analyzes the associated record samples and time patterns of each device to provide control strategies that are more in line with user habits for these scenarios. For example, in "conference mode", the lights in the conference room are adjusted to a moderate brightness, the curtains are automatically closed, and the air conditioning temperature is slightly reduced.

[0121] Real-time environmental monitoring: The system collects environmental information of the office area, such as temperature, humidity, etc., in real time. According to external weather conditions and indoor environmental data, the system automatically adjusts the air-conditioning temperature and wind speed to ensure the comfort of the office environment.

[0122] Response to abnormal situations: If equipment fails or overtemperature is detected, the system immediately triggers emergency instructions, such as shutting down the faulty equipment or starting the backup air conditioner, and sends the alarm information to the management terminal through the notification function to ensure the safety of the office area.

[0123] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A method for controlling an intelligent device, characterized in that: The following steps are involved: Step S1: The user inputs the size information of the workspace, matches it in the workspace template library, and obtains the usage scenario model of the smart device. The usage scenario model is composed of the complete workspace of the user using the smart device. The usage scenario model includes the function identification of each area in the workspace, environmental information and smart device information; Step S2: The user adds multiple smart devices to the usage scenario model through the control terminal, and improves the smart device information of the usage scenario model. The smart device establishes a connection with the control system through a wireless network. The control system obtains parameter information of the smart device through the model information of the smart device, and uniquely marks each smart device. Step S3: define multiple control scenarios on the control terminal, each scenario includes a group of smart devices and corresponding control instructions, and establish a unique control tag for each control instruction of each smart device; the user obtains multiple required control scenarios by adjusting the preset control scenario template; obtain an associated record sample set, each associated record sample corresponds to a control instruction generated by a smart device, the associated record sample is composed of a control tag and a record time point, the associated record sample set includes associated record samples generated by all smart devices over time, and the associated record sample set is sent to the control scenario optimization model for calculation to obtain a linkage control strategy and a habitual control strategy, and the linkage control strategy and the habitual control strategy are used as the control scenario optimization strategy and displayed on the control terminal, and the user updates the control scenario by selecting and determining the control scenario optimization strategy; the control scenario optimization model uses a sliding window to perform statistics on the associated record samples, performs record filtering and association filtering on the statistical results to obtain statistical records, and obtains the linkage control strategy according to the sorting results of the statistical records; Calculate the regularity value of the associated record samples, and obtain the habit control strategy according to the sorting of the regularity values; The control scenario optimization model includes the following structure: The data recording layer is used to record the control instructions of each smart device and convert the records into the format of associated record samples. The associated record samples are composed of control tags and control time. The associated record samples are connected in chronological order to obtain an associated record sample set. The association statistics layer is used to obtain statistical records of association record sample combinations that meet the requirements in the association record sample set, and obtains them by sliding window statistics; it includes the following steps: Step A1: Perform the first round of iteration, set the sliding window to 2 bits, use the sliding window to count the adjacent association record sample combinations in the association record sample one by one, obtain the statistical results, perform record filtering and association filtering on the statistical results, and record the filtered statistical results into statistical records, wherein the statistical records are sets of sets, each set is the statistical result of one round of iteration, and the statistical results are a data string of combination-number-data position of each association record sample combination; Step A101: Filter the associated record sample combination of the first iteration; calculate the time difference between the associated record samples corresponding to the first and second positions in the sliding window. If the time difference exceeds the dynamic association threshold C, do not count the associated record sample combination in the current sliding window; The calculation formula of dynamic correlation threshold C is: , where C0 represents the initial threshold, n represents the number of iterations of the sliding window, the number of iterations n in the first round of iterations is recorded as 0, and the value of n is increased by 1 in each round of iterations, C M Indicates the maximum threshold adjustment value; Step A102: Send the statistical results after record filtering to the correlation calculation layer to calculate the correlation, and perform correlation filtering on the statistical results, and retain the correlation record sample combinations whose correlation meets the standard in the statistical results, where the correlation of the correlation record sample combination is greater than a set threshold; Step A103: Record the set corresponding to the statistical result after the association filtering into the statistical record, completing the first round of sliding window iteration; Step A2: perform the next iteration, use the associated record sample combination recorded in the last sliding window as the first position of the sliding window, find the position of the associated record sample combination through the data position in the last round of statistical results, and count the adjacent n+2 associated record sample combinations in the associated record samples one by one, perform record filtering and association filtering on the statistical results, record the set corresponding to the filtered statistical results into the statistical record, and complete this round of iteration; Step A3: repeat the iterative process in step A2 until there is no associated record sample combination with a satisfactory association degree in the iteration; The correlation calculation layer is used to calculate the correlation L of the associated record sample combination in the statistical results, the correlation L of the i-th associated record sample combination i The calculation formula is: , where N i is the number of times the i-th associated record sample combination appears in the statistical record, T represents the number of days, M is the number of all associated record sample combinations in this iteration, represents the proportion of the i-th associated record combination, j represents the j-th record of the i-th associated record combination, t represents the control time corresponding to the j-th record, d() is the Euclidean distance calculation function, and d(j,t) represents the time similarity of the i-th associated record combination; A and B are the adjustment weights of the proportion and time similarity of the associated record combination, respectively. The initial weight is set to A=B=1 / 2, and A+B=1 is satisfied; The regularity value calculation layer is used to calculate the regularity value of the associated record samples. It is obtained by counting the number of associated record samples in units of days and months. The calculation formula is: , where P d and P m Represent the regularity value of the day and the regularity value of the month, P d By counting the number of associated record samples occurring in the same time period in units of days, i The ratio of the total number of days T is obtained, P m By counting the number of related record samples occurring on the same day in units of months, m di and the total number of months T m The maximum value of P is 1; The strategy summary layer is used to obtain the associated record sample combination and the associated record sample with regular values ​​in the statistical record, and sort them respectively, taking the associated record sample combination with the top in the sorting as the linkage control strategy output, and taking the associated record sample with the top in the sorting as the habit control strategy output; Step S4: acquiring the status information of each smart device, the environment information of the workspace and the weather information of the location in real time, and displaying the current status information of the smart device and the environment information of the workspace through the control terminal; using the status information of each smart device, the environment information of the workspace and the weather information of the location as adjustment control sample data, and sending them into the real-time control model for calculation to obtain adjustment instructions; the real-time control model is trained by historical data to obtain the correlation between the control adjustment sample data and the adjustment instructions; calculating the fitness R of the control instruction and the adjustment instruction, comparing the fitness of the control instruction and the adjustment instruction, and selecting the instruction with the largest fitness to control the smart device; The fitness calculation includes: When there are control instructions and adjustment instructions to control the same device, the adaptability R between the instruction and the device is calculated. The adaptability R between the instruction and the standard state of the environmental information of the current device in its working area is calculated. The adaptability R is the inverse of the distance difference between the standard state, the current state and the state corresponding to the instruction in the device state vector matrix. The smaller the distance difference, the higher the adaptability, and the corresponding instruction is closer to the standard state. The standard state is searched in the device environment table. Step S5: Perform abnormality detection on each smart device according to the status information of the smart device and the environmental information of the workspace to obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the smart device, and send an alarm notification to the control terminal; when performing emergency control, automatically save the working status of the smart device before the emergency control, and the saved record is used to restore the working status of the smart device; the user remotely views and controls the status of the smart device on the control terminal through manual control or voice interaction.

2. A smart device control method according to claim 1, characterized in that: Obtaining the usage scenario model includes: Construction of workspace template library: Get the type of workspace, and build multiple preset workspace templates based on the type. The workspace template is the construction parameter of the use scenario model, which is displayed through a two-dimensional spatial top view. In the template, the workspace is divided into multiple areas based on the physical partition of the space and the recognition of artificial areas. A functional identifier is added to each area. The functional identifier is not unique, and the same area has multiple functional identifiers. The preset workspace template is added to the workspace template library, and the workspace templates are classified by type in the workspace template library; Matching workspace template: The user selects the required workspace category on the control terminal and enters the size information of the workspace. The size information is used to match the workspace template with the closest size in the workspace template library. Parameter adjustment of the workspace template: adjust the matched workspace template according to the input size information to obtain the workspace template of the size required by the user; Build a usage scenario model: Automatically build a usage scenario model based on the final workspace template. The usage scenario model is a three-dimensional space model that includes segmented areas. It is displayed on the control terminal through a 3D model and a top view of the model.

3. A smart device control method according to claim 1, characterized in that: Improve the usage scenario model including: The user adds a smart device on the control terminal by selecting the smart device category or smart device model under the category in the smart device list, and dragging it to the location in the usage scenario model to add it. If the category of the smart device is selected for addition, the model needs to be selected after the addition is completed; or directly click the location in the usage scenario model to add a smart device, and enter the identification information of the smart device to match the model; After completing the addition of smart devices, obtain the parameter information of the smart devices according to their models, calculate the spatial parameters of the corresponding area of ​​each smart device in the usage scenario model, compare the spatial parameters of the area with the parameter information of the smart devices, and perform the comparison through the standard device parameter standard spatial parameter table. If the comparison result deviation exceeds the set threshold, a prompt will be given on the control terminal; The complete usage scenario model includes the function identification, environmental information and smart device information of each area in the workspace. The environmental information is obtained through weather information and sensors. In the usage scenario model, the function identification, environmental information and smart device information are displayed by selecting environmental parameters or through the area list.

4. A smart device control method according to claim 1, characterized in that: Control scenarios include: There are three types of control scenarios: specific trigger scenarios, time status scenarios, and custom scenarios. Specific trigger scenarios are triggered by specific events; time status scenarios are triggered by time, specifically distinguished by month, day, and hour; custom scenarios are customized by users; the three types of control scenarios are compatible. If there is a situation of controlling the same smart device, the adaptability of the control command is calculated, and the command with higher adaptability is selected to control the smart device; The user selects a preset control scene template through the control terminal. The template includes a set of preset smart devices and corresponding control instructions. The user chooses to directly use the preset template or make adjustments based on the preset template.

5. A smart device control method according to claim 1, characterized in that: The real-time control model includes: By acquiring historical data, which includes the status information of smart devices, the environment information of the workspace, and the weather information of the location, and recording the control instructions as classification labels in the historical data, a real-time control data set is constructed through the historical data, and the data set is divided into a training set, a validation set, and a test set; The real-time control training set and verification set are sent to the classification model based on pre-training, and the control instructions are used as the classification target for training to obtain the initial real-time control model; The real-time control test set is sent to the initial real-time control model for accuracy evaluation. If the accuracy is within the set range, the model is output as a real-time control model. Otherwise, the hyperparameters of the initial real-time control model are adjusted and the model is retrained until the accuracy meets the standard.

6. A smart device control method according to claim 1, characterized in that: Anomaly detection includes: Anomaly detection uses a multi-dimensional anomaly detection model, which comprehensively considers the status information of smart devices, the environmental data of the workspace, and the user's historical operation records for anomaly detection. The anomaly detection results include three categories: device operation anomalies, environmental anomalies, and multi-device linkage anomalies. According to different anomaly types, corresponding emergency instructions are generated, and the working status of smart devices is adjusted in real time.

7. An intelligent device control system, the system applying an intelligent device control method according to any one of claims 1 to 6, comprising the following modules: The usage scenario model acquisition module includes a preset template construction unit, a template matching unit, a model generation unit and a model improvement unit. The preset template construction unit is used to build and add preset workspace templates, including background addition and foreground addition. The templates added in the background are displayed to all users by the administrator, and the templates added in the foreground are performed by the user and are only displayed to the user. The user recommends the template added in the foreground, and the template is added to the background after being reviewed and approved by the administrator; the template matching unit is used to match in the workspace template library, and the size information and category of the workspace need to be input; the model generation unit is used to automatically adjust the parameters of the matched workspace template according to the parameters input by the user, and generate a three-dimensional usage scenario model. This unit manually adjusts the workspace template by the user; the model improvement unit is used to add smart devices and mark the regional functions to the usage scenario model, and obtain a usage scenario model including the functional identification of each area in the workspace, environmental information and smart device information; The control scenario construction module includes a scenario construction unit and a scenario management unit. The scenario construction unit is used to obtain multiple required control scenarios by adjusting the preset control scenario template; the scenario management unit is used to adjust the control scenario and select the control scenario optimization strategy output by the control scenario optimization module for automatic adjustment; The control scenario optimization module is composed of a control scenario optimization model, and is used to send the associated record sample set into the control scenario optimization model for calculation, perform statistics on the associated record samples using sliding window statistics, perform record filtering and associated filtering on the statistical results to obtain statistical records, and obtain the linkage control strategy according to the sorting results of the statistical records; Calculate the regularity value of the associated record samples, and obtain the habit control strategy according to the regularity value sorting; use the linkage control strategy and the habit control strategy as the control scenario optimization strategy, and display them on the control terminal; The real-time control module includes an information acquisition unit, an adjustment control unit and a control verification unit. The information acquisition unit is used to acquire the status information of each smart device, the environmental information of the workspace and the weather information of the location in real time. The adjustment control unit is composed of a real-time control model, which is used to use the status information of each smart device, the environmental information of the workspace and the weather information of the location as adjustment control sample data, and send them into the real-time control model for calculation to obtain adjustment instructions; the control verification unit is used to calculate the fitness of the control instruction and the adjustment instruction, compare the fitness of the control instruction and the adjustment instruction, and select the instruction with the large fitness to control the smart device; The abnormality detection module includes an emergency control unit and a record recovery unit. The emergency control unit is used to perform abnormality detection on each intelligent device according to the status information of the intelligent device and the environmental information of the workspace, obtain the abnormality detection result, obtain the emergency instruction according to the abnormality detection result, use the emergency instruction to perform emergency control on the intelligent device, and send an alarm notification to the control terminal; The record recovery unit is used to automatically save the working state of the intelligent device before the emergency control when the emergency control is performed, and the saved record is used to restore the working state of the intelligent device; Visual interface module, used for visual display of the system, including the main interface, model acquisition interface, usage scenario interface and instruction adjustment interface; The main interface is used to provide users with overview information of the control workspace; the model acquisition interface is used to provide users with an operation interface for model acquisition; The usage scenario interface is used to provide users with an interface for adjusting usage scenarios; the instruction adjustment interface is used to provide users with an interface for adjusting instructions.

Images