Edge intelligent terminal interaction management method, device and equipment and storage medium

Abstract

The invention discloses an edge intelligent terminal interaction management method and device, equipment and a storage medium, and relates to the technical field of edge intelligent computing, and the edge intelligent terminal interaction management method comprises the steps: obtaining the use habit and preference setting of a user for an intelligent terminal, and the operation state and environment information of the intelligent terminal; inputting the use habit, the preference setting, the operation state and the environment information into a target energy consumption prediction model to obtain an energy consumption prediction value; adjusting a control strategy for the intelligent terminal according to the energy consumption predicted value to obtain a target control strategy; generating an equipment control instruction according to the target control strategy; and the equipment control instruction is sent to the intelligent terminal, so that the intelligent terminal operates according to the equipment control instruction, and interaction management is completed. The operation state of the intelligent terminal can be adjusted according to the energy consumption, the automation and intelligence level of the system is enhanced, and the energy use efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of edge intelligent computing technology, and in particular to edge intelligent terminal interaction management methods, devices, equipment and storage media. Background Art

[0002] With the widespread application of smart terminals, especially in edge computing and IoT scenarios, energy efficiency management of smart devices has become an urgent problem to be solved. Smart terminals are usually limited by computing resources, storage capacity and energy consumption. How to dynamically adjust energy consumption according to the operating status and load of the terminal, optimize battery life and extend device operation time is an important challenge in the design and management of smart devices. Effective energy management can not only improve the performance of terminal devices, but also reduce energy consumption, promote green computing and sustainable development.

[0003] At present, it is common to save energy by adjusting the processor performance of smart terminals. Specifically, the device dynamically adjusts the operating frequency and voltage of the processor according to the load size. When the load is low, the device will reduce the frequency and voltage to reduce energy consumption; when the load increases, the device will increase the processor performance to ensure the responsiveness and stability of the system. In addition, many devices also support different working modes, such as power saving mode, sleep mode and high-performance mode. The system will automatically switch modes according to the task requirements and device status to optimize energy efficiency.

[0004] Although existing methods have optimized the energy efficiency of smart terminals to a certain extent, there are still some limitations. Existing energy consumption adjustment technologies are often based on static load forecasting and simple resource scheduling strategies, and fail to fully consider the changes in energy consumption under different tasks and environmental conditions. In addition, these methods have weak real-time dynamic adjustment capabilities in complex scenarios and fail to finely manage the system's resources and energy consumption according to actual operating conditions. Especially in edge computing environments, the computing power and energy supply of devices vary greatly. Therefore, how to adjust the operating status of smart terminals based on energy consumption has become an urgent problem to be solved.

[0005] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0006] The purpose of this application is to provide an edge intelligent terminal interaction management method, device, equipment and storage medium, aiming to solve the technical problem of how to adjust the operating status of the intelligent terminal according to energy consumption.

[0007] To achieve the above objectives, the present application proposes an edge intelligent terminal interaction management method, the method comprising:

[0008] Obtaining the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal;

[0009] Inputting the usage habits, the preference settings, the operating status, and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value;

[0010] Adjust the control strategy of the smart terminal according to the energy consumption prediction value to obtain a target control strategy;

[0011] generating device control instructions according to the target control strategy;

[0012] The device control instruction is sent to the smart terminal, so that the smart terminal operates according to the device control instruction to complete the interactive management.

[0013] In one embodiment, the step of sending the device control instruction to the smart terminal so that the smart terminal operates according to the device control instruction and completing the interactive management step further includes:

[0014] Acquire the first current running state and current environment information of the smart terminal corresponding to each edge node and the current usage habits and current preference settings of the user for the smart terminal corresponding to each edge node;

[0015] Inputting the current usage habits, the current preference settings, the first current operating state, and the current environmental information into the target energy consumption prediction model corresponding to each edge node to obtain the power demand of each edge node;

[0016] The parameters of the target energy consumption prediction model corresponding to each edge node and the power demand of the edge node are sent to the cloud application platform, so that the cloud application platform builds a global power demand prediction model according to the parameters and the power demand of the edge node, and adjusts the power distribution of the virtual power grid through the global power demand prediction model.

[0017] In one embodiment, before the step of inputting the usage habits, the preference settings, the operating status, and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value, the step further includes:

[0018] Obtaining historical energy consumption data, historical user usage habits, historical user preference settings, historical operating parameters of the smart terminal, and historical environmental information;

[0019] Standardizing the historical energy consumption data, the historical user usage habits, the historical user preference settings, the historical operating parameters, and the historical environmental information to obtain initial data;

[0020] Convert the initial data into a time series format according to the timestamp to obtain target data;

[0021] Constructing an initial energy consumption prediction model according to the target data;

[0022] Dividing the target data into a training set and a test set;

[0023] Obtaining a reference energy consumption prediction model according to the training set and the initial energy consumption prediction model;

[0024] Inputting the test set into the reference energy consumption prediction model to obtain a predicted value;

[0025] Calculate a mean square error based on the historical energy consumption data and the predicted value;

[0026] When the mean square error is greater than the preset error threshold, return to the step of obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model until the mean square error is less than the preset error threshold, and use the reference energy consumption prediction model as the target energy consumption prediction model.

[0027] In one embodiment, the step of constructing an initial energy consumption prediction model according to the target data includes:

[0028] Performing a unit root test on the target data to obtain a test result;

[0029] When the test result is not stable, the target data is differentiated and the number of differences is recorded, and the step of performing a unit root test on the target data to obtain a test result is returned until the test result is stable;

[0030] Calculate the autocorrelation and partial autocorrelation of the target data using a statistical analysis library;

[0031] Draw an autocorrelation graph and a partial autocorrelation graph according to the autocorrelation and the partial autocorrelation;

[0032] Obtaining the number of autoregressive terms and the number of sliding average terms according to the attenuation points of the autocorrelation graph and the partial autocorrelation graph;

[0033] An initial energy consumption prediction model is constructed according to the number of differences, the number of autoregressive terms, and the number of sliding average terms.

[0034] In one embodiment, the step of obtaining a reference energy consumption prediction model according to the training set and the initial energy consumption prediction model includes:

[0035] Calculate the gradient of the order of the initial energy consumption prediction model according to the training set;

[0036] The search direction is obtained according to the inverse of the Hessian matrix and the gradient calculation;

[0037] Perform a linear search according to the search direction until a preset number of iterations is reached to obtain an autoregressive coefficient, a differential coefficient, and a sliding average coefficient;

[0038] A reference energy consumption prediction model is obtained according to the initial energy consumption prediction model, the autoregressive coefficient, the differential coefficient and the sliding average coefficient.

[0039] In one embodiment, the step of adjusting the control strategy for the smart terminal according to the energy consumption prediction value to obtain a target control strategy includes:

[0040] Obtaining the operating state parameters and control strategy parameters of the smart terminal, the operating state parameters including the on state, off state, standby state, abnormal state and energy-saving state, and the control strategy parameters including the load adjustment strategy, temperature control strategy, device start-stop control strategy and device mode switching strategy;

[0041] Establishing an initial expected utility table according to the on state, the off state, the standby state, the abnormal state, the energy-saving state, the operating state parameters, the load adjustment strategy, the temperature control strategy, the device start-stop control strategy, and the device mode switching strategy;

[0042] Obtaining a target expected utility table according to the operating state, the initial expected utility table and a preset exploration strategy;

[0043] A target control strategy is obtained according to the energy consumption forecast value and the target expected utility table.

[0044] In one embodiment, the step of obtaining a target expected utility table according to the operating state, the initial expected utility table and a preset exploration strategy includes:

[0045] Determining an initial control strategy according to the operating state, the initial expected utility table, and a preset exploration strategy;

[0046] Executing the initial control strategy, and obtaining a second current operating state, an instant reward, and an energy consumption efficiency of the smart terminal;

[0047] updating the initial expected utility table according to the second current operating state, the instant reward, and the energy consumption efficiency to obtain a reference expected utility table;

[0048] Calculating the rate of change between the initial expected utility table and the reference expected utility table;

[0049] When the change rate is greater than or equal to the preset change rate threshold, return to the step of determining the initial control strategy based on the operating state, the initial expected utility table and the preset exploration strategy until the change rate is less than the preset change rate threshold to obtain the target expected utility table.

[0050] In addition, to achieve the above objectives, the present application also proposes an edge intelligent terminal interaction management device, the device comprising:

[0051] A data acquisition module, used to acquire the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal;

[0052] An energy consumption prediction module, used for inputting the usage habits, the preference settings, the operating status and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value;

[0053] A control adjustment module, used to adjust the control strategy of the smart terminal according to the energy consumption prediction value to obtain a target control strategy;

[0054] An instruction generation module, used to generate device control instructions according to the target control strategy;

[0055] The interactive module is used to send the device control instruction to the smart terminal so that the smart terminal operates according to the device control instruction to complete interactive management.

[0056] In addition, to achieve the above-mentioned purpose, the present application also proposes an edge intelligent terminal interaction management device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the edge intelligent terminal interaction management method as described above.

[0057] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the edge intelligent terminal interaction management method described above are implemented.

[0058] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the edge intelligent terminal interaction management method as described above.

[0059] One or more technical solutions proposed in this application have at least the following technical effects:

[0060] Obtain the user's usage habits and preference settings for the smart terminal, as well as the operating status and environmental information of the smart terminal; input the usage habits, preference settings, operating status and environmental information into the target energy consumption prediction model to obtain the energy consumption prediction value; adjust the control strategy for the smart terminal according to the energy consumption prediction value to obtain the target control strategy; generate device control instructions according to the target control strategy; send the device control instructions to the smart terminal so that the smart terminal operates according to the device control instructions to complete interactive management. First, the system collects the user's usage habits and preference settings, as well as the operating status and environmental information of the smart terminal through sensors and user interfaces. These data include the device's working mode, the temperature and brightness set by the user, as well as the device's CPU usage, memory usage, and the temperature and humidity of the environment. The purpose of this is to gain an in-depth understanding of user behavior and device status, and provide accurate input data for subsequent energy consumption prediction. Then, the system inputs these data into the target energy consumption prediction model and obtains the energy consumption prediction value through algorithm processing. This step uses machine learning technology to predict future energy consumption trends based on historical data and current status, helping the system make energy-saving optimization decisions in advance. Then, the system adjusts the control strategy of the smart terminal according to the predicted energy consumption value to form a target control strategy. This step involves optimizing the operating parameters of the equipment according to the prediction results, such as automatically reducing the operating intensity of the equipment when a high energy consumption period is predicted, or increasing the operating load of the equipment when the energy consumption is low, in order to achieve the purpose of energy saving. Next, the system generates specific equipment control instructions according to the target control strategy. These instructions specify in detail how the equipment should adjust its operating state, such as adjusting the temperature set point or changing the working mode, to achieve the predetermined energy saving target. Finally, the system sends these equipment control instructions to the smart terminal, and the smart terminal operates according to the instructions to complete the interactive management. This step ensures that the equipment can respond to the control instructions in real time, realize intelligent energy management, improve energy efficiency, reduce costs, and enhance user experience. Overall, the system can adjust the operating state of the smart terminal according to energy consumption, which not only improves the efficiency of energy use and the accuracy of prediction, but also enhances the automation and intelligence level of the system, providing users with a more comfortable and energy-saving living environment, while also saving costs and reducing energy waste for enterprises and users. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0062] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0063] Figure 1 A schematic diagram of the interactive management system architecture provided for the edge intelligent terminal interactive management method of this application;

[0064] Figure 2 A flowchart of the first embodiment of the edge intelligent terminal interaction management method of the present application is provided;

[0065] Figure 3 A flow chart of the second embodiment of the edge intelligent terminal interaction management method of the present application;

[0066] Figure 4 This is a schematic diagram of the module structure of the edge intelligent terminal interaction management device according to an embodiment of the present application;

[0067] Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the edge intelligent terminal interaction management method in the embodiment of the present application.

[0068] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0069] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0070] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0071] The widespread use of smart terminals in edge computing and the Internet of Things brings challenges to energy efficiency management, as they are often limited by computing, storage resources, and energy consumption. Dynamically adjusting energy consumption to optimize battery life and extend operating time is key to design and management. Common energy-saving methods are to dynamically adjust processor performance according to load, including operating frequency and voltage, and automatically switch between different operating modes. However, these methods have limitations, such as being based on static load prediction and simple resource scheduling, and lacking the ability to make real-time dynamic adjustments in complex scenarios, especially in edge computing environments with large differences in computing and energy supply.

[0072] The main solution of the embodiment of the present application is to collect the user's usage habits, preference settings, and the operating status and environmental information of the smart terminal through sensors and user interfaces, such as the device's working mode, temperature, brightness, CPU usage, etc., in order to gain an in-depth understanding of user behavior and device status, and provide accurate data for energy consumption forecasting. Subsequently, machine learning technology is used to process these data, predict future energy consumption trends, and make energy-saving optimization decisions in advance. Based on the predicted values, the system adjusts the control strategy of the smart terminal, such as automatically reducing the operating intensity during high energy consumption periods and increasing the load during low energy consumption periods to achieve energy saving. Based on this, the system generates specific device control instructions, such as adjusting the temperature setting or changing the working mode, and sends them to the smart terminal for execution, completing the entire interactive management process.

[0073] It should be noted that the execution subject of the embodiment of the present application may be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, an interactive management system, etc. The following takes the interactive management system as an example to illustrate this embodiment and the following embodiments.

[0074] The interactive management system is a multi-layered architecture, including an edge platform layer and an edge node layer. The edge platform layer is connected to the cloud application platform layer, and the edge node layer is connected to multiple smart terminals. The cloud application platform layer is at the top level and is responsible for deploying global models and intelligent algorithms. It uses a microservice architecture to ensure the flexibility of the system. It communicates with the edge platform layer through APIs and is responsible for model updates and intelligent decision-making. The edge platform layer is located between the cloud and edge nodes, using edge computing nodes for local data storage and computing resource configuration. It supports data filtering and preprocessing, reduces data transmission volume, and improves response speed. The edge node layer deploys lightweight applications on each edge node, which can receive instructions from the cloud and send control commands to terminal devices. This layer is also responsible for implementing containerization technology and task scheduling to ensure efficient communication and resource utilization between nodes. Smart terminals are located at the bottom of the architecture, interacting directly with users or other systems, receiving control commands from the edge node layer, and feeding back status information.

[0075] In this system, the introduction of serverless computing architecture enables computing resources to be dynamically adjusted according to terminal loads, achieve automatic expansion, and optimize costs and efficiency. The edge node layer deploys lightweight applications on each node, receives cloud instructions and sends control commands to terminal devices. At the same time, containerization technology is implemented to encapsulate different functional modules into containers, and container orchestration tools such as Kubernetes are used to automatically manage and distribute loads to ensure task scheduling and efficient communication and resource utilization between nodes. The integration of deep learning is reflected in multiple levels, including the introduction of reinforcement learning algorithms in edge nodes, dynamic optimization of control strategies based on real-time data, and improved decision-making intelligence. Federated learning technology allows each edge node to perform local model training and regularly upload model parameters to the cloud for aggregation to update the global model. In addition, lightweight deep learning models are developed and optimized using frameworks such as TensorFlow Lite or PyTorch Mobile to ensure that they adapt to the memory and computing limitations of edge devices. Finally, in order to improve the user experience, user-friendly mobile applications and web front-ends are developed, and intuitive interfaces are designed for users to monitor device status, receive alarms, and control devices. Overall, this interactive management system achieves intelligent, automated and efficient equipment management and control by integrating deep learning technologies such as reinforcement learning and federated learning, as well as new technologies such as serverless architecture.

[0076] Please refer to Figure 1 , Figure 1 The schematic diagram of the interactive management system architecture provided for the edge intelligent terminal interactive management method of this application shows a layered interactive management system architecture, from the top cloud application platform layer to the edge platform layer, and then to the edge node layer, and finally to the bottom intelligent terminal layer. The cloud application platform layer is at the top of the architecture, responsible for deploying global models and intelligent algorithms, using microservice architecture to ensure the flexibility of the system, and communicating with the edge platform layer through API. The edge platform layer is located between the cloud and edge nodes, using edge computing nodes for local data storage and computing resource configuration, supporting data filtering and preprocessing to reduce data transmission and improve response speed. The edge node layer deploys lightweight applications on each edge node, which can receive instructions from the cloud and send control commands to terminal devices, while implementing containerization technology and task scheduling to ensure efficient communication and resource utilization between nodes. The intelligent terminal layer includes intelligent terminal 1, intelligent terminal 2, and intelligent terminal 3, etc., which interact directly with users or other systems, receive control commands from the edge node layer, and feedback status information. Through this layered design, the entire architecture realizes intelligent, automated and efficient equipment management and control, optimizes energy efficiency, and improves the responsiveness and stability of the system.

[0077] Based on this, the embodiment of the present application provides an edge intelligent terminal interaction management method, referring to Figure 2 , Figure 2 This is a flow chart of the first embodiment of the edge intelligent terminal interaction management method of the present application.

[0078] In this embodiment, the edge intelligent terminal interaction management method includes steps S10 to S50:

[0079] Step S10, obtaining the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal;

[0080] It should be noted that smart terminals refer to a type of embedded computer system devices that have the same architecture framework and the same hardware and software structure as embedded devices. These terminal devices place special emphasis on intelligence and networking, and can interact with users or other systems, receive control commands, and feedback status information. Examples of smart terminals include but are not limited to: (1) Smart home devices: such as smart air conditioners, smart light bulbs, smart sockets, etc., which can be connected to the Internet to achieve remote control and energy management. (2) Industrial automation equipment: robots and automation instruments used on production lines, which can achieve equipment control and monitoring through smart terminals to improve production efficiency. (3) Medical and health equipment: such as smart blood pressure monitors, smart blood glucose meters, smart electrocardiographs, etc., these devices can obtain patients' physiological parameters and provide doctors with a basis for diagnosis. (4) Logistics management equipment: such as smart warehousing systems and smart logistics vehicles, which can achieve rapid sorting, tracking and management of goods through smart terminals. (5) Mobile smart terminals: including smartphones and tablets, which have independent operating systems, can install software and games, and achieve wireless network access through mobile communication networks. Usage habits and preference settings refer to the personal preferences and behavior patterns of users when interacting with smart terminals. This may include user adjustments to device brightness, volume, notification settings, or specific operating habits of users when using applications, such as temperature settings for smart air conditioners, brightness adjustments for smart light bulbs, and timer switches for smart sockets. By obtaining this information, the system can better understand user needs, thereby providing personalized services and optimizing user experience. Operating status and environmental information involve the current working status of the smart terminal, including hardware status indicators such as processor load, memory usage, battery power, temperature, and the environmental conditions of the device, such as light intensity, sound level, temperature, and humidity. This information is critical to the system because it directly affects the energy efficiency management and performance optimization of the device. For example, if a smart terminal detects that the battery is low, the system may automatically adjust the processor performance to extend battery life.

[0081] It is understandable that, first, the system collects the usage data of the smart terminal through the built-in sensors and user interface, including the user's specific operation records and setting preferences for the smart terminal, such as the temperature adjustment of the smart air conditioner, the brightness selection of the smart light bulb, etc., in order to capture the user's personalized needs and usage patterns. Then, the system collects the device's operating status and environmental information, such as battery power, CPU load, temperature and humidity, by monitoring the hardware status of the smart terminal and the external environmental sensors. These data are essential for evaluating the health of the device and optimizing energy consumption. Finally, by analyzing the collected data, the system can dynamically adjust the working parameters of the smart terminal, such as processor frequency and voltage, and automatically switch the working mode to adapt to different usage scenarios and environmental conditions. The effect of this is to optimize the energy efficiency of the device, extend battery life, improve device performance and user experience, while reducing energy consumption and promoting green computing and sustainable development.

[0082] As an example, before the step of obtaining the user's usage habits and preference settings for the smart terminal and the operating status and environmental information of the smart terminal, it also includes: determining the smart terminal registration mechanism and the smart terminal authentication method; receiving the registration request sent by the smart terminal, including information such as device ID, device type and user information; detecting the validity and integrity of the registration request to obtain a detection result (such as device ID format check); when the detection result is passed, accepting the registration request and generating a unique identifier (such as UUID) for the smart terminal; generating device registration information according to the device ID, the device type, the user information and the unique identifier, and sending it to the smart terminal; receiving the authentication request sent by the smart terminal, including device ID and device registration information; verifying the device ID and device registration information to obtain a verification result; when the verification result is passed, generating an access token (such as JWT), and sending the access token to the smart terminal, the token containing the identity information and validity period of the smart terminal; configuring an access control policy to only receive connection requests containing valid access tokens.

[0083] Smart terminal registration mechanism and smart terminal authentication method refer to a series of processes and methods to ensure the legitimacy of the identity of smart terminals and allow them to access resources when they join a network or system. A registration request refers to a request sent by a smart terminal to the system, which contains information such as device ID, device type and user information, which is the basic data for the system to identify and record devices. Device ID is the unique identifier of a smart terminal, such as IMEI, MEID or UUID, which is used to distinguish different devices. Device type refers to the type of smart terminal, such as smart phones, smart air conditioners, etc., while user information includes account information related to users. Validity and integrity refer to the verification of information in the registration request to ensure that the information has not been tampered with during transmission and that all necessary information is included. The test result refers to the result of the system's verification of the registration request, which may be a pass or fail status. A unique identifier (such as UUID) refers to a unique code generated by the system for each registered smart terminal, which is used to uniquely identify the device in the system. Device registration information refers to comprehensive information including device ID, device type, user information and unique identifier, which is used to record and manage devices in the system. An authentication request refers to a request sent by a smart terminal to the system after registration in order to obtain access rights, which contains the device ID and device registration information. The verification result refers to the result of the system's verification of the authentication request, which determines whether the device is allowed to access resources. An access token (such as JWT) refers to a security token sent by the system to a smart terminal after successful authentication, which contains the identity information and validity period of the smart terminal and is used to verify the identity of the device in subsequent communications. Identity information refers to the information contained in the access token that can identify the smart terminal or user, and the validity period refers to the validity period limit of the access token, after which the token will no longer be valid. Access control policy refers to a series of rules set by the system based on the identity and permissions of the smart terminal, and only receives connection requests containing valid access tokens to ensure secure access to resources. Such a policy can prevent unauthorized access and potential security threats, ensuring that only legitimate devices can connect to the system.

[0084] First, the registration mechanism and authentication method of the smart terminal need to be determined to ensure that the device can safely join the network and be recognized by the system. Next, the system receives the registration request sent by the smart terminal, which contains key data such as device ID, device type, and user information. The system will detect the validity and integrity of this information, such as checking whether the format of the device ID is correct. This step is to ensure the accuracy and security of the information. If the test result shows that the registration request is valid, the system will accept the request and generate a unique identifier (UUID) for the smart terminal, which is unique in the entire network. Then, the system will generate device registration information based on the device ID, device type, user information, and unique identifier, and send it back to the smart terminal. Subsequently, the system receives the authentication request sent by the smart terminal, which contains the device ID and device registration information. The system will verify this information to confirm the identity of the device. If the verification result is passed, the system will generate an access token (such as JWT), which contains the identity information and validity period of the smart terminal, ensuring the security and timeliness of access, and send this token to the smart terminal. Finally, the system configures access control policies to allow only connection requests that contain valid access tokens. This ensures that only authenticated devices can access network resources, thereby protecting system security and data integrity.

[0085] Step S20, inputting the usage habits, the preference settings, the operating status and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value;

[0086] It should be noted that the target energy consumption prediction model refers to a model built using machine learning or data mining technology, which can predict the energy consumption of the smart terminal based on input data (such as user usage habits, preference settings, operating status of the smart terminal, and environmental information). This model can be based on a variety of algorithms, including linear regression, support vector machines, neural networks, etc., to analyze and predict energy consumption data. In this embodiment, the target energy consumption prediction model is established based on the autoregressive integrated moving average model (ARIMA), and the general form is:

[0087]

[0088] Among them, Y t is the value of the time series data at time t, φ i is the coefficient of the autoregressive term, θ j is the coefficient of the moving average term, and B is the backoff operator, so that B i Y t =Y t-i , d is the order of difference, used to make the time series stationary, ∈ tis the white noise error term, p is the number of autoregressive terms, and q is the number of moving average terms.

[0089] The energy consumption prediction value refers to the prediction result calculated by the model based on the input data, that is, the estimated energy consumption of the smart terminal in the future. This prediction value can help users or system administrators understand the energy consumption trend of the equipment, so as to carry out energy management and optimization. In this way, energy utilization efficiency can be improved, costs can be reduced, and green and sustainable development can be supported.

[0090] It is understandable that inputting the acquired usage habits, preference settings, operating status, and environmental information into the target energy consumption prediction model means using these data as input variables and analyzing them through the algorithms and learning mechanisms within the model. This model is specially designed to process and predict energy consumption. It can identify complex relationships between input data and predict the energy consumption trends of smart terminals based on these relationships. The model can be built based on machine learning or deep learning technologies, such as neural networks, support vector machines, or deep learning algorithms. The model processes these inputs, comprehensively considers the impact of various factors on energy consumption, and finally outputs an energy consumption prediction value, which is the model's quantitative prediction of the energy consumption of smart terminals in a specific time period in the future, which can be used to guide the implementation of energy-saving measures and optimize the energy efficiency of equipment.

[0091] As an example, before the step of inputting the usage habits, the preference settings, the operating status and the environmental information into the target energy consumption prediction model to obtain the energy consumption prediction value, it also includes: obtaining historical energy consumption data, historical user usage habits, historical user preference settings, historical operating parameters of the smart terminal and historical environmental information; standardizing the historical energy consumption data, the historical user usage habits, the historical user preference settings, the historical operating parameters and the historical environmental information to obtain initial data; converting the initial data into a time series format according to a timestamp to obtain target data; constructing an initial energy consumption prediction model based on the target data; dividing the target data into a training set and a test set; obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model; inputting the test set into the reference energy consumption prediction model to obtain a prediction value; calculating a mean square error based on the historical energy consumption data and the prediction value; when the mean square error is greater than a preset error threshold, returning to the step of obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model until the mean square error is less than the preset error threshold, and using the reference energy consumption prediction model as the target energy consumption prediction model.

[0092] Historical energy consumption data refers to the record of energy consumption of smart terminals in the past period of time, such as energy consumption data in the past 24 hours, a week, or a month. This data can help analyze energy consumption trends and patterns and provide a basis for predicting future energy consumption. Historical user usage habits refer to the habitual behaviors formed when users interact with smart terminals, such as when users usually use the device and which functions they use. These habits will affect the energy consumption pattern of the device. Historical user preference settings refer to the personalized settings of users for smart terminals, such as screen brightness, volume, temperature preferences, etc. These settings directly or indirectly affect the energy consumption of the device. Historical operating parameters refer to the operating status parameters of smart terminals in historical periods, including CPU and memory usage, storage space usage, etc. These parameters reflect the workload of the device and have a direct impact on energy consumption. Historical environmental information refers to the record of environmental conditions in which the smart terminal is located, such as indoor temperature, humidity, light intensity, etc. Environmental factors may affect the energy consumption of the device, especially in devices that require environmental adjustment (such as air conditioning and lighting). Standardization is a data processing process that aims to convert data from different sources and different dimensions into a unified metric for comparison and analysis, such as converting temperature from Celsius to Fahrenheit, or scaling data of different dimensions to a range of 0 to 1. Initial data refers to a data set that has been standardized and is ready to be used to build and train a prediction model. A timestamp refers to an identifier that indicates a specific point in time, usually including year, month, day, hour, minute and second. A timestamp is used to identify the specific moment of data recording. A time series format refers to a data format that is arranged in chronological order, which can show the trend of data changes over time. In energy consumption forecasting, time series data can reveal the changes in energy consumption at different times of the day. Target data refers to the data obtained after arranging the initial data in chronological order. These data will be used to build and train the energy consumption prediction model. The initial energy consumption prediction model refers to the preliminary version of the energy consumption prediction model built based on the target data. It may not be accurate enough and needs to be optimized through training and testing. The target data is divided into two parts: the training set is used to train the model, and the test set is used to evaluate the performance of the model. The training set helps the model learn the patterns in the data, and the test set is used to verify the prediction ability of the model. The reference energy consumption prediction model refers to the model trained using the training set, which serves as the benchmark for subsequent optimization and comparison. The predicted value refers to the result obtained by predicting the test set data using the reference energy consumption prediction model. These values ​​will be compared with the actual energy consumption data. The mean square error (MSE) refers to a statistical indicator that measures the accuracy of model predictions. The calculation method is to take the square of the difference between the predicted value and the actual value, and then find the average. The smaller the MSE, the more accurate the model's prediction. The preset error threshold refers to an error value set during the model training process as the acceptance standard for model performance. If the MSE of the model is lower than this threshold, the model is considered good enough to be used for actual energy consumption prediction.For example, if the preset error threshold is set to 0.01, this means that the average squared difference between the energy consumption predicted by the model and the actual energy consumption should not exceed 0.01. If the MSE of the model exceeds this value, it is necessary to return and retrain the model until the MSE is lower than the threshold, at which point the model is considered to be an effective energy consumption prediction model.

[0093] The interactive management system first collects the historical energy consumption data of smart terminals, the user's historical usage habits, historical preference settings, historical operating parameters of the equipment, and environmental information in order to accumulate enough data to train an accurate energy consumption prediction model. Then, the system standardizes this data to eliminate the influence of different dimensions and scales so that the data can be compared and analyzed under the same standard, which can improve the efficiency of model training and the accuracy of prediction. The formula for data standardization is:

[0094]

[0095] Among them, Z is the standardized value, X is the original data value, μ is the mean of the data set, and σ is the standard deviation of the data set.

[0096] Next, the system converts the standardized data into a time series format in chronological order. This can capture the trend of energy consumption changes over time and provide richer information for building an energy consumption prediction model. Subsequently, the system uses these time series data to build an initial energy consumption prediction model, which will be used for subsequent training and testing. In order to evaluate the performance of the model, the system divides the time series data into a training set and a test set. The training set is used to train the model, and the test set is used to verify the predictive ability of the model. The system uses the training set to train the initial model to obtain a reference energy consumption prediction model, and then inputs the test set into this model to obtain the predicted value. The system compares these predicted values ​​with the actual historical energy consumption data and calculates the mean square error. This error indicator is used to measure the accuracy of the model's prediction. The calculation formula for the mean square error (MSE) is:

[0097]

[0098] Among them, Y i is the actual value, is the predicted value and n is the number of data points.

[0099] If the mean square error is greater than the preset error threshold, it means that the prediction accuracy of the model is not enough, and the system will return to the training step and continue to optimize the model until the mean square error is lower than the preset threshold, thus ensuring that the final energy consumption prediction model has high prediction accuracy and reliability. Through these steps, the system can build a model that can accurately predict the energy consumption of smart terminals, providing a scientific basis for the energy consumption management and optimization of smart terminals, thereby improving energy efficiency and reducing energy consumption costs.

[0100] As an example, the step of constructing an initial energy consumption prediction model based on the target data includes: performing a unit root test on the target data to obtain a test result; when the test result is not stable, performing a difference on the target data and recording the number of differences, and returning to the step of performing a unit root test on the target data to obtain a test result until the test result is stable; using a statistical analysis library to calculate the autocorrelation and partial autocorrelation of the target data; drawing an autocorrelation graph and a partial autocorrelation graph based on the autocorrelation and the partial autocorrelation; obtaining the number of autoregressive terms and the number of sliding average terms based on the attenuation points of the autocorrelation graph and the partial autocorrelation graph; and constructing an initial energy consumption prediction model based on the number of differences, the number of autoregressive terms, and the number of sliding average terms.

[0101] The unit root test is a statistical test used to determine whether time series data is stationary. The statistical properties of a stationary time series (such as mean and variance) do not change over time, while a non-stationary time series exhibits a trend or seasonal pattern. The purpose of the unit root test is to detect whether there is a unit root in the time series, that is, whether the series exhibits the characteristics of a random walk over time. Commonly used unit root tests include the Dickey-Fuller (DF) test and the augmented Dickey-Fuller (ADF) test. The statistical calculation formula for the ADF test is:

[0102]

[0103] in, is the autoregressive coefficient of the time series, is the standard error of the coefficient.

[0104] The test result of the unit root test is usually a statistic and a p-value. If the p-value is less than the preset significance level (for example, 0.05), the null hypothesis is rejected and the series is considered to be stationary. If the p-value is greater than the significance level, the null hypothesis cannot be rejected and the series is considered to be non-stationary. Non-stationary: If the statistical characteristics of a time series (such as mean, variance, or autocorrelation) change over time, then the series is non-stationary. Non-stationarity may be caused by trends, seasonality, or other forms of systematic changes. Difference is a method of converting a non-stationary time series into a stationary series. It eliminates the trend and seasonal components in the series by calculating the differences between consecutive observations. For example, the first-order difference is to calculate the difference between the value at each time point and the value at the previous time point. The formula for first-order differences is:

[0105] Δy t =y t -y t-1

[0106] Among them, y t is the value of the time series at time t, y t-1 is the value of the time series at time t-1.

[0107] The number of differences refers to the number of times a time series needs to be differentiated in order to make it stationary. Different time series may require different numbers of differences, from first order to multiple orders. Statistical analysis libraries refer to software libraries that provide statistical analysis functions, such as NumPy, SciPy, Pandas, and statsmodels in Python. These libraries provide tools for performing statistical tasks such as unit root tests, autocorrelation, and partial autocorrelation analysis. Autocorrelation is a measure of the correlation between values ​​of a time series at different time lags, while partial autocorrelation is the correlation between values ​​of a time series at a specific lag given other lag values. These measures help identify dependency structures in time series. The formula for calculating the autocorrelation coefficient is:

[0108]

[0109] Among them, E represents the expected value, μ represents the mean, and σ represents the standard deviation.

[0110] The autocorrelation plot and partial autocorrelation plot respectively show the change of autocorrelation and partial autocorrelation of the time series with the number of lags. They are useful tools for identifying time series model parameters (such as the number of autoregressive terms and the number of moving average terms). In the autocorrelation plot and partial autocorrelation plot, the decay point refers to the point where the correlation changes from significant to insignificant. The lag after this point can be considered to contribute little to the model, so it can be used to determine the order of the model. The number of autoregressive terms (p) refers to the number of lag terms included in the autoregressive model, while the number of moving average terms (q) refers to the number of lag terms included in the moving average model. These parameters determine the complexity and fitting ability of the model. The initial energy consumption prediction model is the initial version of the time series prediction model constructed based on the above analysis steps. It is usually an autoregressive integrated moving average (ARIMA) model, including autoregressive terms, number of differences and moving average terms. This model will be used for subsequent optimization and prediction to accurately predict energy consumption values.

[0111] In the process of building the initial energy consumption prediction model, the interactive management system first performs a unit root test on the target data. This step is to detect whether the time series maintains a constant mean and variance over time by calculating statistics, because only a stable time series can be effectively predicted. If the test results show that the data is not stable, the system will perform differential processing on the data, that is, calculate the difference between consecutive observations. This is done to remove the trend and seasonality in the data and make the data stable, because the data after differential is easier to analyze and predict. The system will record the number of differences and repeat the unit root test until the data is stable. This process ensures the stability of the model and the accuracy of the prediction. Next, the system uses the statistical analysis library to calculate the autocorrelation and partial autocorrelation of the target data. These calculations reveal the correlation of data points at different time lags and help identify potential patterns in the time series. Then, the system draws autocorrelation and partial autocorrelation plots based on these correlation measures. These charts intuitively show how the correlation of the data changes with time lags. By observing the decay points in these charts, the system determines the number of autoregressive terms and the number of sliding average terms. These parameters represent the degree to which past values ​​in the model affect the current value and the degree to which the current error affects the prediction. Finally, the system combines the number of differences, the number of autoregressive terms, and the number of moving average terms to build an initial energy consumption prediction model. This model can capture the dynamic changes of time series data and provide a solid foundation for predicting future energy consumption, making energy consumption management more accurate and efficient.

[0112] As an example, the step of obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model includes: calculating the gradient of the order of the initial energy consumption prediction model based on the training set; obtaining a search direction based on the inverse of the Hessian matrix and the gradient calculation; performing a linear search according to the search direction until a preset number of iterations is reached to obtain an autoregressive coefficient, a differential coefficient and a sliding average coefficient; and obtaining a reference energy consumption prediction model based on the initial energy consumption prediction model, the autoregressive coefficient, the differential coefficient and the sliding average coefficient.

[0113] Order: In statistics, order usually refers to the highest degree of the variables in the model. For example, in polynomial regression, the order is the degree of the polynomial. In the energy consumption prediction model, the order refers to the number of lag terms in the time series in the model, that is, how many past time points the model considers to predict future values. Gradient: In machine learning, the gradient refers to the derivative of the loss function with respect to the model parameters. It indicates the direction in which the loss function rises fastest in the parameter space. During the optimization process, the gradient is used to guide parameter updates to minimize the loss function. The gradient calculation formula is:

[0114]

[0115] Among them, J is the loss function and θ represents the model parameters.

[0116] Inverse of the Hessian matrix: The Hessian matrix is ​​a matrix composed of the second-order derivatives of the loss function with respect to the model parameters. Its inverse can be used to calculate the search direction for parameter updates in optimization problems. In particular, in Newton's method, the inverse of the Hessian matrix provides a way to measure the mutual influence between parameters. Search direction: In the optimization algorithm, the search direction refers to the direction of movement from one point to another when looking for the minimum value in the parameter space. This direction is based on the information of the gradient and the Hessian matrix, and the purpose is to find the fastest path to reduce the loss function value. The calculation formula for the search direction in the Newton method is:

[0117]

[0118] Where H is the Hessian matrix, is the gradient, H- 1 is the inverse of the Hessian matrix, and d is the search direction.

[0119] Linear search is a search method for finding the optimal parameters. It determines the parameter value that minimizes the loss function by linear scanning along a given search direction. This process involves gradually adjusting the parameters in the search direction until a preset number of iterations is reached or the minimum value is found. This process is usually determined using a line search algorithm, such as the Armijo rule or other optimization algorithms. The formula can be expressed as:

[0120] α=arg min α J(θ-αd)

[0121] Among them, α is the step size, θ is the current parameter value, and d is the search direction.

[0122] Preset number of iterations: This is a parameter set during the optimization process. It defines the maximum number of iterations the algorithm performs before stopping. Preset number of iterations helps control computational costs and prevent overfitting. Autoregressive coefficient, difference coefficient, and sliding average coefficient: These coefficients are key parameters in time series analysis. The autoregressive coefficient represents the relationship between the current value and its past value. The difference coefficient is related to the difference of the time series and is used to remove trends and seasonality, while the sliding average coefficient represents the relationship between the current value and the past error terms. The reference energy consumption prediction model is an optimized model based on the above parameters and the initial energy consumption prediction model. It improves the prediction accuracy of the model by combining the autoregressive coefficient, difference coefficient and sliding average coefficient, as well as the data in the training set. The formula for constructing a reference energy consumption prediction model (such as the ARIMA model) is:

[0123]

[0124] Among them, Y t is the value of the time series data at time t, φ i is the autoregressive coefficient, θ j is the moving average coefficient, B is the backward operator, d is the order of difference, ∈t is the error term. This model combines autoregressive terms, difference terms and moving average terms to predict future energy consumption values.

[0125] In the process of building the reference energy consumption prediction model, the interactive management system first evaluates the parameters of the initial energy consumption prediction model through the training set. The specific method is to calculate the gradient of the model parameters, that is, the rate of change of the loss function for each parameter. This is done to find out how to adjust the parameters to reduce the prediction error as quickly as possible. Then, the system calculates the inverse of the Hessian matrix, which contains the second-order derivative information between the parameters and is used together with the gradient to determine the search direction for parameter updates. This direction indicates the fastest path to reduce losses in the parameter space. Next, the system performs a linear search along this search direction, that is, gradually trying different parameter values, and iterating and optimizing until the preset number of iterations is reached. This ensures that the model fully explores the parameter space and finds the optimal solution. In this process, the system updates the autoregressive coefficients, differential coefficients, and sliding average coefficients, which determine how the model uses historical data to predict future energy consumption. Finally, combining these optimized coefficients with the initial model, the system constructs a reference energy consumption prediction model, which can more accurately capture the dynamic changes of time series data, improve the accuracy and reliability of predictions, and thus provide more effective decision support for energy management.

[0126] Step S30, adjusting the control strategy for the smart terminal according to the energy consumption prediction value to obtain a target control strategy;

[0127] It should be noted that the control strategy in the application of smart terminals refers to a set of rules or algorithms used to manage and adjust the behavior of terminal devices to achieve specific goals, such as energy saving, optimizing performance or meeting user needs. The target control strategy is a control strategy adjusted according to the predicted energy consumption value. It guides the operation of smart terminals more specifically and accurately to achieve energy consumption optimization and efficiency improvement.

[0128] It can be understood that the system first receives the predicted values ​​output by the energy consumption prediction model, which reflect the expected energy consumption of the smart terminal under current and future conditions. Then, the system analyzes these predicted values ​​to determine whether there are any situations that exceed the energy consumption budget or are inefficient. Based on these analyses, the system adjusts the control strategy, such as adjusting the device's operating mode, modifying user preference settings, optimizing task scheduling, or adjusting device parameters to reduce energy consumption or improve energy efficiency. These adjustments are designed to achieve more efficient energy use and ensure that the device operates in the best state while meeting user needs and system performance requirements. Ultimately, the control strategy formed by these adjustments is called the target control strategy, which directly guides the actual operation of the smart terminal to achieve the predetermined energy consumption goals and performance standards. In this way, the smart terminal can respond to environmental changes and user needs more intelligently, achieving energy saving and optimized operation.

[0129] Step S40, generating a device control instruction according to the target control strategy;

[0130] It should be noted that device control instructions refer to a series of specific operation commands generated by the system, which are used to guide the smart terminal on how to make actual adjustments and operations according to the target control strategy. These instructions may include but are not limited to: (1) Adjusting operating parameters: such as modifying the operating mode of the smart terminal, such as switching the air conditioner from cooling mode to energy-saving mode, or adjusting the temperature set point. (2) Optimizing task scheduling: such as postponing or prioritizing certain tasks to reduce the load during peak energy consumption periods. (3) Modifying user preferences: such as automatically reducing screen brightness or closing unused applications to reduce energy consumption. (4) Device parameter adjustment: such as adjusting the voltage and frequency of the device to optimize energy efficiency. (5) Starting and stopping devices: shutting down the device when it is not needed, or starting the device when demand increases. (6) Maintenance and early warning: such as performing maintenance or issuing early warnings when it is predicted that the device may overheat or performance deteriorate.

[0131] It is understandable that first, the system analyzes the target control strategies, which may be based on the results of the energy consumption prediction model and aim to reduce energy consumption, improve equipment efficiency or meet specific performance standards. The system then converts these strategies into specific equipment operation commands, such as adjusting the temperature setting of smart air conditioners, changing the brightness of smart light bulbs, optimizing the switching time of smart sockets, or adjusting the load distribution of servers. These control instructions will be sent to the corresponding smart terminals to instruct them to perform specific actions to achieve predetermined energy efficiency and performance goals. In this way, smart terminals can automatically respond to control instructions and optimize their operating status, thereby achieving the effects of energy saving, improving efficiency and meeting user needs.

[0132] Step S50, sending the device control instruction to the smart terminal, so that the smart terminal operates according to the device control instruction to complete the interactive management.

[0133] It can be understood that the system first determines the specific actions that need to be performed based on the target control strategy, such as lowering the temperature setting of a smart thermostat or adjusting the brightness level of a smart light bulb. The system then translates these actions into machine-readable control instructions, which usually involves writing specific code or using a predefined command format. These instructions are encapsulated into data packets and sent to the smart terminal via wireless signals or wired networks. The receiver or interface built into the smart terminal captures these data packets and decodes them to perform the corresponding actions. The purpose of this is to enable the smart terminal to automatically respond to external instructions and adjust its operating state to achieve energy efficiency optimization and performance management. In this way, the smart terminal can dynamically adjust its behavior based on real-time data and prediction results, thereby reducing energy waste, improving energy efficiency, and possibly extending the service life of the equipment. Ultimately, this interactive management can achieve a more intelligent and automated home or industrial environment, improving user comfort and operational convenience.

[0134] As an example, the device control instruction is sent to the smart terminal so that the smart terminal operates according to the device control instruction. After completing the interactive management step, it also includes: obtaining the first current operating status and current environmental information of the smart terminal corresponding to each edge node and the user's current usage habits and current preference settings for the smart terminal corresponding to each edge node; inputting the current usage habits, the current preference settings, the first current operating status and the current environmental information into the target energy consumption prediction model corresponding to each edge node to obtain the power demand of each edge node; sending the parameters of the target energy consumption prediction model corresponding to each edge node and the power demand of the edge node to the cloud application platform, so that the cloud application platform constructs a global power demand prediction model based on the parameters and the power demand of the edge node, and adjusts the power distribution of the virtual power grid through the global power demand prediction model.

[0135] Edge nodes refer to nodes located at the edge of the network, close to users or data sources in distributed computing and network architectures, and are used to provide fast data transmission, processing, computing and storage services. Their role is to bring data and computing resources as close to data sources and end users as possible, thereby reducing latency, improving response speed, and reducing the load on central servers. First, the current operating status and current environmental information refer to the operating status and environmental conditions of the smart terminal at a certain moment, such as the device's CPU usage, memory usage, battery power, and environmental temperature and humidity. Current usage habits and current preference settings refer to habitual behaviors and personalized settings formed when users interact with smart terminals, such as when users are used to using devices, preferred brightness settings, volume levels, etc. Edge node power demand refers to the power consumption demand of the node predicted in the future based on the operating status, environmental information, user usage habits and preference settings of the edge node. The parameters of the target energy consumption prediction model corresponding to each edge node refer to the parameters used to construct the energy consumption prediction model. These parameters may include autoregressive coefficients, differential coefficients, sliding average coefficients, etc., which together define how the model predicts future energy consumption based on historical data. The cloud application platform refers to a service platform located in the cloud layer. It consists of multiple high-performance servers and storage devices, has powerful computing and storage capabilities, can perform complex computing tasks, and manage and schedule edge nodes. The global power demand forecasting model refers to the overall power demand forecasting model constructed by the cloud application platform based on the power demand forecasting model parameters and power demand of each edge node, which is used to predict the power consumption demand of the entire system. The virtual grid is a power coordination management system that realizes the aggregation and coordinated optimization of distributed energy resources such as distributed power sources, energy storage systems, and controllable loads through advanced information and communication technologies and software systems. The virtual grid does not generate electricity directly, but participates in the power market and power grid operation by aggregating and optimizing power resources. Power allocation refers to the allocation and management of power resources by the virtual grid according to the power grid demand and power market conditions to achieve the optimization of power supply and the stable operation of the power grid.

[0136] First, the system collects real-time data from smart terminals from each edge node, including the operating status of the device, such as CPU and memory usage, and environmental information, such as indoor temperature and humidity, because these parameters directly affect the energy consumption of the device. At the same time, the system also records the user's usage habits and preferences, such as brightness and temperature preferences, because these can be used for personalized energy consumption prediction. Then, the system inputs this data into the energy consumption prediction model of each edge node. These models can predict the power demand of each node in the future based on the input data, which can help the system plan and adjust resources in advance. Next, the system uploads the energy consumption prediction model parameters and power demand of each node to the cloud application platform. The cloud platform will integrate these data to build a global power demand prediction model, which can predict the power consumption trend of the entire system, thereby providing a basis for the load management and power dispatch of the power grid. Finally, the cloud platform adjusts the power distribution of the virtual power grid according to the output of the global power demand prediction model, optimizes the allocation of power resources, and ensures the balance and stability of the supply and demand of the power grid. The effect of this is to improve the efficiency of energy use and the accuracy of prediction, reduce energy waste, and also enhance the power grid's ability to respond to peak loads, and improve the intelligence level of energy management and dispatch of the entire system.

[0137] This embodiment provides an edge intelligent terminal interactive management method, which obtains the user's usage habits and preference settings for the intelligent terminal, as well as the operating status and environmental information of the intelligent terminal; inputs the usage habits, the preference settings, the operating status and the environmental information into the target energy consumption prediction model to obtain the energy consumption prediction value; adjusts the control strategy for the intelligent terminal according to the energy consumption prediction value to obtain the target control strategy; generates a device control instruction according to the target control strategy; sends the device control instruction to the intelligent terminal so that the intelligent terminal operates according to the device control instruction to complete the interactive management. First, the system collects the user's usage habits and preference settings, as well as the operating status and environmental information of the intelligent terminal through sensors and user interfaces. These data include the device's working mode, the temperature and brightness set by the user, as well as the device's CPU usage, memory usage, and the temperature and humidity of the environment. The purpose of this is to gain an in-depth understanding of user behavior and device status, and to provide accurate input data for subsequent energy consumption prediction. Then, the system inputs these data into the target energy consumption prediction model and obtains the energy consumption prediction value through algorithm processing. This step uses machine learning technology to predict future energy consumption trends based on historical data and current status, helping the system make energy-saving optimization decisions in advance. Then, the system adjusts the control strategy of the smart terminal according to the predicted energy consumption value to form a target control strategy. This step involves optimizing the operating parameters of the equipment according to the prediction results, such as automatically reducing the operating intensity of the equipment when a high energy consumption period is predicted, or increasing the operating load of the equipment when the energy consumption is low, in order to achieve the purpose of energy saving. Next, the system generates specific equipment control instructions according to the target control strategy. These instructions specify in detail how the equipment should adjust its operating state, such as adjusting the temperature set point or changing the working mode, to achieve the predetermined energy saving target. Finally, the system sends these equipment control instructions to the smart terminal, and the smart terminal operates according to the instructions to complete the interactive management. This step ensures that the equipment can respond to the control instructions in real time, realize intelligent energy management, improve energy efficiency, reduce costs, and enhance user experience. Overall, the system can adjust the operating state of the smart terminal according to energy consumption, which not only enhances the automation and intelligence level of the system, but also improves the efficiency of energy use, provides users with a more comfortable and energy-saving living environment, and also saves costs and reduces energy waste for enterprises and users.

[0138] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can refer to the above introduction, and will not be repeated later. Figure 3 , Figure 3 This is a flow chart of the second embodiment of the edge intelligent terminal interaction management method of the present application. Step S30 of the edge intelligent terminal interaction management method includes steps S31 to S34:

[0139] Step S31, obtaining the operating state parameters and control strategy parameters of the smart terminal, the operating state parameters including the on state, off state, standby state, abnormal state and energy-saving state, and the control strategy parameters including the load adjustment strategy, temperature control strategy, device start-stop control strategy and device mode switching strategy;

[0140] It should be noted that the operating status parameters cover the parameters of the smart terminal in different working states, including: (1) On state: refers to the smart terminal being turned on and in active working mode, able to execute tasks and respond to user operations. (2) Off state: the smart terminal is completely turned off, does not consume power, and does not perform any operations. (3) Standby state: the smart terminal is turned on but does not perform any operations, is in low-power waiting mode, and is ready to respond to user operations at any time. (4) Abnormal state: the smart terminal fails or is abnormal and cannot work normally, and may require manual intervention or automatic repair. (5) Energy-saving state: the smart terminal enters a special mode to reduce energy consumption, which may reduce performance but reduce energy consumption. Control strategy parameters involve how to manage and adjust the operation of smart terminals to achieve different goals, including: (1) Load adjustment strategy: dynamically adjust resource allocation and task processing according to the system load to optimize performance and energy consumption. For example, reduce device performance to save energy when the load is low, and improve performance to meet demand when the load is high. (2) Temperature control strategy: Temperature control of smart terminals, such as reducing the load or turning on the cooling system when the temperature is too high to protect the device and save energy. This can be achieved through algorithms such as PID control, fuzzy control, and adaptive control to keep the device running at the optimal temperature. (3) Device start-stop control strategy: Strategies for deciding when to start or stop device operation, such as shutting down the device to save energy when demand is low, or quickly starting the device to meet demand when demand is high. (4) Device mode switching strategy: Strategies for smart terminals to switch between different operating modes based on environmental changes or user needs, such as switching to energy-saving mode at night or switching to high-performance mode when the load is high.

[0141] It is understandable that the operating status parameters of the device need to be obtained first, which includes monitoring whether the device is in the on state (the device is operating normally and responding to user operations), the off state (the device is completely turned off and does not consume energy), the standby state (the device is waiting for user operations and is in low-energy mode), the abnormal state (the device encounters a fault and needs maintenance or repair) and the energy-saving state (the device reduces performance to reduce energy consumption). At the same time, it is also necessary to collect control strategy parameters, which involve load adjustment strategies (dynamically adjusting resource allocation according to device load to balance performance and energy consumption), temperature control strategies (monitoring and adjusting device temperature to prevent overheating and optimize energy consumption), device start-stop control strategies (automatically starting or shutting down devices based on demand to save energy) and device mode switching strategies (automatically switching the working mode of the device according to the usage scenario or time, such as switching from high-performance mode to energy-saving mode). The acquisition of these parameters is to monitor and adjust the operation of smart terminals in real time to ensure that the device optimizes energy use, improves efficiency, and extends the life of the device while meeting user needs. By precisely controlling these parameters, smart terminals can respond to environmental changes and user needs more intelligently, achieving the dual goals of energy saving and improving user experience.

[0142] Step S32, establishing an initial expected utility table according to the on state, the off state, the standby state, the abnormal state, the energy-saving state, the operating state parameters, the load adjustment strategy, the temperature control strategy, the device start-stop control strategy, and the device mode switching strategy;

[0143] It should be noted that the initial expected utility table refers to a data structure used in the Q-learning algorithm, which stores the expected utility (or Q value) of the agent taking a specific action in a specific state. When this table is initialized, it usually contains all possible combinations of states and actions, but the Q value may be random or zero at the beginning. As the algorithm runs, the agent will update these values ​​based on actual experience to reflect the long-term utility of each state-action pair. Specifically, the general process of establishing the expected utility table is as follows: (1) Create a Q table: First, create a table for each possible state and action. The rows of this table represent states, the columns represent possible actions, and each cell corresponds to a state-action pair, and its value represents the expected utility of taking the action in that state. (2) Initialize Q values: At the beginning of the Q-learning algorithm, all Q values ​​are usually initialized to zero or a small random number. This is because without experience, we have no information to distinguish the utilities of different state-action pairs. (3) Update Q-value: As the agent interacts in the environment, it updates the Q-value based on the actual reward and the expected maximum utility of the next state. This process involves the core update formula of Q-learning, which is to adjust the Q-value of the current state-action pair based on the Q-value of the current state, the immediate reward obtained, and the maximum expected utility of the next state. (4) Select action: The agent selects an action based on the current Q-table, usually using the ε-greedy strategy, that is, selecting the action with the highest current Q-value with a certain probability and randomly selecting an action with a certain probability to balance exploration and exploitation. (5) Repeat process: The agent repeatedly selects actions, observes results, and updates Q-values ​​until the Q-table converges. At this time, the agent can select the optimal action strategy based on the Q-table.

[0144] It can be understood that first, the system needs to define all possible states and actions. States include the on state, off state, standby state, abnormal state and energy-saving state of the smart terminal, which reflect the different working modes and conditions of the device. Actions include load regulation strategy, temperature control strategy, device start-stop control strategy and device mode switching strategy, which guide how the device responds to different operating conditions and user needs. Next, the system creates a table for all state-action pairs, namely the Q table. The rows of this table represent states, the columns represent actions, and each cell corresponds to the Q value of a state-action pair, namely the expected utility. Initially, these Q values ​​are usually set to zero or a small random number because there is no prior experience to estimate the utility. The system collects the operating state parameters of the smart terminal, such as the CPU usage rate of the device, the ambient temperature, etc. These parameters provide real-time data on device performance and environmental conditions. At the same time, the system also integrates control strategy parameters, such as load regulation and temperature control strategies, as well as device start-stop and mode switching strategies. As the smart terminal operates in the actual environment, the system updates the Q value according to the Q-learning algorithm. Specifically, the system adjusts the Q value of the current state-action pair based on the immediate reward obtained by the device after taking a specific action and the maximum expected utility of the next state. This process involves the core update formula of Q-learning, that is, the update of the Q value is based on the current Q value, the reward obtained, and the maximum Q value of the next state. The system controls the actions of the smart terminal according to the current Q table, usually using the ε-greedy strategy, that is, selecting the action with the highest current Q value with a certain probability, and randomly selecting actions with a certain probability to balance exploration and utilization. Through this process, the system is able to learn to control the smart terminal to take the best action under different states to maximize the long-term cumulative reward. This not only improves energy efficiency and reduces energy consumption, but also may extend the service life of the equipment. At the same time, this automated energy management and cost savings also bring higher efficiency and better performance to the operation of smart terminals.

[0145] Step S33, obtaining a target expected utility table according to the operating state, the initial expected utility table and a preset exploration strategy;

[0146] It should be noted that the exploration strategy in the Q-learning algorithm refers to a method for balancing exploration and exploitation. In the process of Q-learning, the agent needs to choose between the known best action (exploitation) and trying new actions (exploration). In this embodiment, the preset exploration strategy refers to the ε-greedy strategy, where ε is a probability value less than 1, which is used to determine the probability of the agent randomly selecting an action in each time step, and the current optimal action is selected with a probability of 1-ε. The purpose of this is to allow the agent to maintain a balance between exploring the environment to find better strategies and using known information to obtain immediate rewards. This strategy helps the agent find better strategies during the learning process while avoiding falling into local optimal solutions. The target expected utility table is a core component of the Q-learning algorithm, which records the expected utility of the agent on each state-action pair, that is, the Q value. When this table is initialized, the Q value of each state-action pair is usually set to zero or a small random number, indicating that nothing is known about the expected utility of these actions. As the agent interacts and learns in the environment, these Q values ​​are constantly updated based on the agent's immediate rewards and expectations of future rewards. The target expected utility table will eventually contain the agent's optimal strategy learned through learning, that is, the actions that should be taken in each state to maximize the long-term cumulative reward. This table allows the agent to predict the long-term utility of different actions based on historical data and preset exploration strategies, and make decisions accordingly.

[0147] It can be understood that the system first starts the learning process based on the current operating state of the intelligent terminal, such as on, off, standby, abnormal or energy-saving state, and the initial expected utility table, which is a table that records the Q values ​​of all possible state-action pairs, where the Q value represents the expected utility of taking a specific action in a specific state. Initially, these Q values ​​may be randomly assigned or zero, because there is not enough information to determine which action is the best. Then, the system implements a preset exploration strategy, such as the ε-greedy strategy, which determines whether the agent randomly selects an action with a certain probability (ε) to explore new possibilities at each time step, or selects the currently known best action with a higher probability (1-ε) to utilize existing knowledge. In this way, the agent interacts in the environment and updates the Q value based on the actual rewards obtained and the expectation of future rewards. This process is repeated until the Q value converges to form a target expected utility table, which reflects the optimal expected utility of the agent for each state-action pair after long-term learning, and guides the agent to select actions that can maximize long-term cumulative rewards in different states. Ultimately, the target expected utility table enables the agent to make more precise and efficient decisions, optimizing its behavior to adapt to environmental changes.

[0148] As an example, the step of obtaining a target expected utility table according to the operating state, the initial expected utility table and the preset exploration strategy includes: determining an initial control strategy according to the operating state, the initial expected utility table and the preset exploration strategy; executing the initial control strategy, and obtaining a second current operating state, an instant reward and an energy consumption efficiency of the smart terminal; updating the initial expected utility table according to the second current operating state, the instant reward and the energy consumption efficiency to obtain a reference expected utility table; calculating the rate of change of the initial expected utility table and the reference expected utility table; when the rate of change is greater than or equal to a preset rate of change threshold, returning to the step of determining the initial control strategy according to the operating state, the initial expected utility table and the preset exploration strategy until the rate of change is less than the preset rate of change threshold to obtain the target expected utility table.

[0149] The initial control strategy refers to the action plan determined based on the current operating state of the intelligent terminal and the initial expected utility table (Q table). This strategy is formulated at the beginning of the algorithm based on the preset exploration strategy (such as the ε-greedy strategy) to find a balance between exploring new actions and using the best known actions. The second current operating state refers to the new state of the intelligent terminal after executing the initial control strategy, such as changing from standby state to on state. The immediate reward refers to the feedback obtained by the intelligent terminal after performing a specific action, which may be positive (such as energy saving) or negative (such as excessive energy consumption). Energy efficiency refers to the energy consumption performance of the intelligent terminal after performing an action, which is used to evaluate the energy efficiency of the action. The reference expected utility table refers to the new table obtained after updating the initial expected utility table based on the second current operating state, immediate reward and energy efficiency. The Q value of this table more accurately reflects the expected utility of taking different actions under different states. The rate of change refers to the difference ratio between the initial expected utility table and the reference expected utility table, which is used to measure the magnitude of the Q value update. Calculating the rate of change can help the system judge the learning progress and the degree of strategy improvement. The preset change rate threshold is a pre-set value used to determine whether the Q value update is significant enough to decide whether to continue the learning process. If the change rate is greater than or equal to this threshold, it means that the Q value update is significant and the learning process will continue; if it is less than the threshold, it is considered that the Q value has stabilized and the learning process can be stopped. For example, if the preset change rate threshold is set to 0.01, then only when the change rate between the initial expected utility table and the reference expected utility table is greater than or equal to 0.01, the system will consider the update to be significant and continue the exploration and learning process. When the change rate is less than 0.01, the system believes that the Q value is stable enough to stop updating, and the reference expected utility table becomes the target expected utility table.

[0150] The system first analyzes the current operating state of the device, such as whether it is on, off, or in standby mode, and looks at the Q values ​​in the initial expected utility table, which represent the expected effects of taking different actions in different states. Combined with the preset exploration strategy, the system determines an initial control strategy to determine the specific actions that the device should perform next, such as adjusting the temperature or switching modes. This is done to optimize energy consumption while maintaining device performance. The system then executes this initial control strategy and immediately collects the new operating state of the device, immediate rewards (which may be positive feedback for energy saving or negative feedback for high energy consumption), and energy efficiency data after execution. This data provides the actual effect after executing the control strategy, helping the system understand which actions are effective and which need to be improved. The system then uses this new data to update the initial expected utility table to generate a reference expected utility table. The update process involves adjusting the Q value to reflect the newly acquired experience. For example, if an action leads to reduced energy consumption, the Q value of this action will increase, indicating that this action is more likely to be selected in similar states in the future. The system calculates the rate of change between the initial expected utility table and the updated reference expected utility table. This rate of change measures the magnitude of the Q value update and is a key indicator for judging the learning progress. If the rate of change is greater than or equal to the preset rate of change threshold, such as 0.01, the system will consider the Q value update to be significant and need to continue learning to further optimize the control strategy. Therefore, it will return to the step of determining the initial control strategy and repeat the entire learning process. This process will continue until the rate of change is less than the preset rate of change threshold, at which time the Q value is considered to be stable and the learning process can be stopped. Ultimately, the target expected utility table obtained by the system provides the optimal expected utility for taking actions under different states. This table guides the smart terminal to make more accurate decisions in actual operation, optimize energy consumption and improve equipment performance. Through this continuous learning and adjustment, the smart terminal can automatically adapt to environmental changes, improve energy efficiency, reduce operating costs, and possibly extend the service life of the equipment.

[0151] Step S34, obtaining a target control strategy according to the energy consumption prediction value and the target expected utility table.

[0152] It should be noted that the target control strategy refers to a set of optimized control instructions based on the energy consumption prediction value and the target expected utility table. These instructions are intended to guide the intelligent terminal on how to adjust its operating parameters and behaviors to achieve energy consumption optimization and system performance optimization. In short, the target control strategy is to determine the specific actions that the device should take in different situations based on the predicted energy consumption and expected utility to improve energy efficiency and responsiveness. For example, suppose that the smart thermostat predicts that energy consumption will increase significantly after 10 pm in the coming winter by analyzing historical energy consumption data and the temperature preferences set by the user. At this time, the target expected utility table records the expected energy consumption and user comfort utility values ​​under different temperature settings. Based on this information, the system recognizes that lowering the thermostat setting from 22℃ to 20℃ can significantly reduce energy consumption while still keeping the user's comfort within an acceptable range. Therefore, the target control strategy is to automatically adjust the thermostat temperature setting from 22℃ to 20℃ at 10 pm. This strategy is based on energy consumption forecasts and expected utility tables. It takes into account the balance between energy saving and user comfort, and aims to optimize overall energy efficiency while ensuring that the user's comfort experience is not greatly affected. In this way, the smart thermostat can automatically adjust its operation to adapt to environmental changes and user needs, achieving smarter energy management.

[0153] Understandably, the system first analyzes the energy consumption forecast values, which are based on historical data and learning algorithms to predict energy consumption in different time periods in the future. At the same time, the system refers to the target expected utility table, which is a table that records the expected utility (such as energy saving effect and user satisfaction) corresponding to different control actions. Combining these two data, the system evaluates various possible control actions, such as adjusting the operating parameters of the equipment or changing the working mode, to determine which action can maximize energy saving while meeting user needs. Finally, the system selects an action plan, namely the target control strategy, which details how the device should respond under specific conditions to optimize energy consumption and performance, such as reducing the operating intensity of the device when high energy consumption periods are predicted, or improving performance during low energy consumption periods to meet user needs. Such a strategy helps smart terminals make the best decisions in different situations to maximize energy efficiency and save costs.

[0154] This embodiment obtains the operating state parameters and control strategy parameters of the smart terminal, the operating state parameters include the open state, the closed state, the standby state, the abnormal state and the energy-saving state, and the control strategy parameters include the load adjustment strategy, the temperature control strategy, the device start-stop control strategy and the device mode switching strategy; according to the open state, the closed state, the standby state, the abnormal state, the energy-saving state, the operating state parameters, the load adjustment strategy, the temperature control strategy, the device start-stop control strategy and the device mode switching strategy, an initial expected utility table is established; according to the operating state, the initial expected utility table and the preset exploration strategy, a target expected utility table is obtained; according to the energy consumption prediction value and the target expected utility table, a target control strategy is obtained. The system collects the operating state parameters of the device by monitoring the current state of the smart terminal, such as the open, closed, standby, abnormal or energy-saving state, and the control strategy parameters, such as load adjustment, temperature control, device start-stop and mode switching strategy. This is done to fully understand the working state of the device and the executable operation range, and to provide data support for the formulation of effective control strategies. Using the collected operating status and control strategy parameters, the system initializes an expected utility table to record the expected utility (Q value) of each state-action pair. This table provides a starting point for the Q-learning algorithm for the subsequent learning process, enabling the system to begin learning how to select the best action based on the state. The system executes actions based on the current operating state, the initial expected utility table, and the preset exploration strategy (such as the ε-greedy strategy), collects results, and updates the expected utility table. This process enables the system to learn and optimize the expected utility table to reflect the actual best action selection by actually executing actions and observing the results, thereby improving the decision-making quality of the system. Combining the energy consumption forecast value and the updated target expected utility table, the system determines the best control strategy under different energy consumption forecast conditions. This strategy details how the device should respond under specific conditions to optimize energy consumption and performance, such as reducing the operating intensity of the device when a high energy consumption period is predicted, or improving performance during a low energy consumption period to meet user needs. Such a strategy helps smart terminals make the best decisions in different situations, optimizing energy consumption and improving device performance. Overall, the system can automatically adjust its control strategy through continuous learning and optimization to adapt to environmental changes and user needs, achieving the dual goals of energy saving and improving user experience. Through this continuous learning and adjustment, the smart terminal can automatically optimize its operation strategy, improve energy efficiency, reduce operating costs, and potentially extend the service life of the equipment.

[0155] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the edge intelligent terminal interaction management method of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0156] This application also provides an edge intelligent terminal interaction management device, please refer to Figure 4 , the edge intelligent terminal interaction management device includes:

[0157] The data acquisition module 10 is used to acquire the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal;

[0158] The energy consumption prediction module 20 is used to input the usage habits, the preference settings, the operating status and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value;

[0159] A control adjustment module 30, configured to adjust the control strategy for the smart terminal according to the energy consumption prediction value to obtain a target control strategy;

[0160] An instruction generation module 40, used to generate device control instructions according to the target control strategy;

[0161] The interactive module 50 is used to send the device control instruction to the smart terminal, so that the smart terminal operates according to the device control instruction to complete the interactive management.

[0162] In one embodiment, the interaction module 50 is also used to obtain the first current operating state and current environmental information of the smart terminal corresponding to each edge node and the user's current usage habits and current preference settings for the smart terminal corresponding to each edge node; input the current usage habits, the current preference settings, the first current operating state and the current environmental information into the target energy consumption prediction model corresponding to each edge node to obtain the power demand of each edge node; send the parameters of the target energy consumption prediction model corresponding to each edge node and the power demand of the edge node to the cloud application platform, so that the cloud application platform constructs a global power demand prediction model according to the parameters and the power demand of the edge node, and adjusts the power distribution of the virtual power grid through the global power demand prediction model.

[0163] In one embodiment, the energy consumption prediction module 20 is also used to obtain historical energy consumption data, historical user usage habits, historical user preference settings, historical operating parameters of the smart terminal, and historical environmental information; standardize the historical energy consumption data, the historical user usage habits, the historical user preference settings, the historical operating parameters, and the historical environmental information to obtain initial data; convert the initial data into a time series format according to the timestamp to obtain target data; construct an initial energy consumption prediction model based on the target data; divide the target data into a training set and a test set; obtain a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model; input the test set into the reference energy consumption prediction model to obtain a predicted value; calculate a mean square error based on the historical energy consumption data and the predicted value; when the mean square error is greater than a preset error threshold, return to the step of obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model until the mean square error is less than the preset error threshold, and use the reference energy consumption prediction model as the target energy consumption prediction model.

[0164] In one embodiment, the energy consumption prediction module 20 is also used to perform a unit root test on the target data to obtain a test result; when the test result is not stable, the target data is differentiated and the number of differentiations is recorded, and the step of performing a unit root test on the target data to obtain a test result is returned until the test result is stable; the autocorrelation and partial autocorrelation of the target data are calculated using a statistical analysis library; an autocorrelation graph and a partial autocorrelation graph are plotted based on the autocorrelation and the partial autocorrelation; the number of autoregressive terms and the number of sliding average terms are obtained based on the attenuation points of the autocorrelation graph and the partial autocorrelation graph; an initial energy consumption prediction model is constructed based on the number of differentiations, the number of autoregressive terms and the number of sliding average terms.

[0165] In one embodiment, the energy consumption prediction module 20 is also used to calculate the gradient of the order of the initial energy consumption prediction model based on the training set; obtain the search direction based on the inverse of the Hessian matrix and the gradient calculation; perform a linear search according to the search direction until a preset number of iterations is reached to obtain an autoregressive coefficient, a differential coefficient and a sliding average coefficient; and obtain a reference energy consumption prediction model based on the initial energy consumption prediction model, the autoregressive coefficient, the differential coefficient and the sliding average coefficient.

[0166] In one embodiment, the control adjustment module 30 is also used to obtain the operating state parameters and control strategy parameters of the smart terminal, the operating state parameters include an on state, an off state, a standby state, an abnormal state and an energy-saving state, and the control strategy parameters include a load regulation strategy, a temperature control strategy, a device start-stop control strategy and a device mode switching strategy; an initial expected utility table is established according to the on state, the off state, the standby state, the abnormal state, the energy-saving state, the operating state parameters, the load regulation strategy, the temperature control strategy, the device start-stop control strategy and the device mode switching strategy; a target expected utility table is obtained according to the operating state, the initial expected utility table and the preset exploration strategy; a target control strategy is obtained according to the energy consumption prediction value and the target expected utility table.

[0167] In one embodiment, the control adjustment module 30 is also used to determine an initial control strategy based on the operating state, the initial expected utility table and a preset exploration strategy; execute the initial control strategy and obtain a second current operating state, an instant reward and an energy consumption efficiency of the smart terminal; update the initial expected utility table based on the second current operating state, the instant reward and the energy consumption efficiency to obtain a reference expected utility table; calculate the rate of change of the initial expected utility table and the reference expected utility table; when the rate of change is greater than or equal to a preset change rate threshold, return to the step of determining the initial control strategy based on the operating state, the initial expected utility table and the preset exploration strategy until the rate of change is less than the preset change rate threshold to obtain a target expected utility table.

[0168] The edge intelligent terminal interaction management device provided by the present application adopts the edge intelligent terminal interaction management method in the above embodiment, which can solve the technical problem of how to adjust the operating state of the intelligent terminal according to energy consumption. Compared with the prior art, the beneficial effects of the edge intelligent terminal interaction management device provided by the present application are the same as the beneficial effects of the edge intelligent terminal interaction management method provided by the above embodiment, and the other technical features in the edge intelligent terminal interaction management device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0169] The present application provides an edge intelligent terminal interaction management device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the edge intelligent terminal interaction management method in the above-mentioned embodiment one.

[0170] Reference below Figure 5, which shows a schematic diagram of the structure of an edge intelligent terminal interaction management device suitable for implementing the embodiment of the present application. The edge intelligent terminal interaction management device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The edge intelligent terminal interaction management device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0171] like Figure 5 As shown, the edge intelligent terminal interaction management device may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. In RAM1004, various programs and data required for the operation of the edge intelligent terminal interaction management device are also stored. The processing device 1001, ROM1002 and RAM1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 can allow the edge intelligent terminal interaction management device to communicate wirelessly or wired with other devices to exchange data. Although the figure shows an edge intelligent terminal interaction management device with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have alternatively.

[0172] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0173] The edge intelligent terminal interaction management device provided by the present application adopts the edge intelligent terminal interaction management method in the above embodiment, which can solve the technical problem of how to adjust the operating state of the intelligent terminal according to energy consumption. Compared with the prior art, the beneficial effects of the edge intelligent terminal interaction management device provided by the present application are the same as the beneficial effects of the edge intelligent terminal interaction management method provided by the above embodiment, and the other technical features in the edge intelligent terminal interaction management device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0174] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0175] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0176] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the edge intelligent terminal interaction management method in the above-mentioned embodiment.

[0177] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0178] The above-mentioned computer-readable storage medium may be included in the edge intelligent terminal interaction management device; or it may exist independently without being assembled into the edge intelligent terminal interaction management device.

[0179] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the edge intelligent terminal interaction management device, the edge intelligent terminal interaction management device: obtains the user's usage habits and preference settings for the intelligent terminal and the operating status and environmental information of the intelligent terminal; inputs the usage habits, the preference settings, the operating status and the environmental information into the target energy consumption prediction model to obtain an energy consumption prediction value; adjusts the control strategy for the intelligent terminal according to the energy consumption prediction value to obtain a target control strategy; generates a device control instruction according to the target control strategy; and sends the device control instruction to the intelligent terminal so that the intelligent terminal operates according to the device control instruction to complete the interaction management.

[0180] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0181] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0182] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0183] The readable storage medium provided in this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned edge intelligent terminal interaction management method, and can solve the technical problem of how to adjust the operating state of the intelligent terminal according to energy consumption. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the edge intelligent terminal interaction management method provided in the above-mentioned embodiment, and will not be repeated here.

[0184] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the edge intelligent terminal interaction management method as described above.

[0185] The computer program product provided by the present application can solve the technical problem of how to adjust the operating state of the smart terminal according to energy consumption. Compared with the prior art, the beneficial effects of the computer program product provided by the present application are the same as the beneficial effects of the edge smart terminal interaction management method provided by the above-mentioned embodiment, which will not be repeated here. The above is only a partial embodiment of the present application, and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the specification and drawings of this application under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for interactive management of edge intelligent terminals, characterized in that: The method comprises: Obtaining the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal; Inputting the usage habits, the preference settings, the operating status, and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value; Adjust the control strategy of the smart terminal according to the energy consumption prediction value to obtain a target control strategy; generating device control instructions according to the target control strategy; The device control instruction is sent to the smart terminal, so that the smart terminal operates according to the device control instruction to complete the interactive management.

2. The method according to claim 1, characterized in that After the step of sending the device control instruction to the smart terminal so that the smart terminal operates according to the device control instruction and completing the interactive management, the method further includes: Acquire the first current running state and current environment information of the smart terminal corresponding to each edge node and the current usage habits and current preference settings of the user for the smart terminal corresponding to each edge node; Inputting the current usage habits, the current preference settings, the first current operating state, and the current environmental information into the target energy consumption prediction model corresponding to each edge node to obtain the power demand of each edge node; The parameters of the target energy consumption prediction model corresponding to each edge node and the power demand of the edge node are sent to the cloud application platform, so that the cloud application platform builds a global power demand prediction model according to the parameters and the power demand of the edge node, and adjusts the power distribution of the virtual power grid through the global power demand prediction model.

3. The method according to claim 1, characterized in that Before the step of inputting the usage habits, the preference settings, the operating status and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value, the step further includes: Obtaining historical energy consumption data, historical user usage habits, historical user preference settings, historical operating parameters of the smart terminal, and historical environmental information; Standardizing the historical energy consumption data, the historical user usage habits, the historical user preference settings, the historical operating parameters, and the historical environmental information to obtain initial data; Convert the initial data into a time series format according to the timestamp to obtain target data; Constructing an initial energy consumption prediction model according to the target data; Dividing the target data into a training set and a test set; Obtaining a reference energy consumption prediction model according to the training set and the initial energy consumption prediction model; Inputting the test set into the reference energy consumption prediction model to obtain a predicted value; Calculate a mean square error based on the historical energy consumption data and the predicted value; When the mean square error is greater than the preset error threshold, return to the step of obtaining a reference energy consumption prediction model based on the training set and the initial energy consumption prediction model until the mean square error is less than the preset error threshold, and use the reference energy consumption prediction model as the target energy consumption prediction model.

4. The method according to claim 3, characterized in that The step of constructing an initial energy consumption prediction model according to the target data comprises: Performing a unit root test on the target data to obtain a test result; When the test result is not stable, the target data is differentiated and the number of differences is recorded, and the step of performing a unit root test on the target data to obtain a test result is returned until the test result is stable; Calculate the autocorrelation and partial autocorrelation of the target data using a statistical analysis library; Draw an autocorrelation graph and a partial autocorrelation graph according to the autocorrelation and the partial autocorrelation; Obtaining the number of autoregressive terms and the number of sliding average terms according to the attenuation points of the autocorrelation graph and the partial autocorrelation graph; An initial energy consumption prediction model is constructed according to the number of differences, the number of autoregressive terms, and the number of sliding average terms.

5. The method according to claim 3, characterized in that The step of obtaining a reference energy consumption prediction model according to the training set and the initial energy consumption prediction model comprises: Calculate the gradient of the order of the initial energy consumption prediction model according to the training set; The search direction is obtained according to the inverse of the Hessian matrix and the gradient calculation; Perform a linear search according to the search direction until a preset number of iterations is reached to obtain an autoregressive coefficient, a differential coefficient, and a sliding average coefficient; A reference energy consumption prediction model is obtained according to the initial energy consumption prediction model, the autoregressive coefficient, the differential coefficient and the sliding average coefficient.

6. The method according to claim 1, characterized in that The step of adjusting the control strategy for the smart terminal according to the energy consumption prediction value to obtain a target control strategy comprises: Obtaining the operating state parameters and control strategy parameters of the smart terminal, the operating state parameters including the on state, off state, standby state, abnormal state and energy-saving state, and the control strategy parameters including the load adjustment strategy, temperature control strategy, device start-stop control strategy and device mode switching strategy; Establishing an initial expected utility table according to the on state, the off state, the standby state, the abnormal state, the energy-saving state, the operating state parameters, the load adjustment strategy, the temperature control strategy, the device start-stop control strategy, and the device mode switching strategy; Obtaining a target expected utility table according to the operating state, the initial expected utility table and a preset exploration strategy; A target control strategy is obtained according to the energy consumption forecast value and the target expected utility table.

7. The method according to claim 6, characterized in that The step of obtaining a target expected utility table according to the operating state, the initial expected utility table and a preset exploration strategy comprises: Determining an initial control strategy according to the operating state, the initial expected utility table, and a preset exploration strategy; Executing the initial control strategy, and obtaining a second current operating state, an instant reward, and an energy consumption efficiency of the smart terminal; updating the initial expected utility table according to the second current operating state, the instant reward, and the energy consumption efficiency to obtain a reference expected utility table; Calculating the rate of change between the initial expected utility table and the reference expected utility table; When the change rate is greater than or equal to the preset change rate threshold, return to the step of determining the initial control strategy based on the operating state, the initial expected utility table and the preset exploration strategy until the change rate is less than the preset change rate threshold to obtain the target expected utility table.

8. An edge intelligent terminal interaction management device, characterized in that: The device comprises: A data acquisition module, used to acquire the user's usage habits and preference settings for the smart terminal and the operating status and environment information of the smart terminal; An energy consumption prediction module, used for inputting the usage habits, the preference settings, the operating status and the environmental information into a target energy consumption prediction model to obtain an energy consumption prediction value; A control adjustment module, used to adjust the control strategy of the smart terminal according to the energy consumption prediction value to obtain a target control strategy; An instruction generation module, used to generate device control instructions according to the target control strategy; The interactive module is used to send the device control instruction to the smart terminal so that the smart terminal operates according to the device control instruction to complete interactive management.

9. An edge intelligent terminal interaction management device, characterized in that: The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the edge intelligent terminal interaction management method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the edge intelligent terminal interaction management method as described in any one of claims 1 to 7 are implemented.

Images