Ubiquitous resource fusion management method and system
Through the ubiquitous resource convergence management method across the cloud-end, equipment data is collected and analyzed in real time, and related models are established and optimized, the problem that traditional methods are difficult to manage multiple device resources is solved, and efficient and automated resource management is achieved.
Patent Information
- Application Number
- CN202411645713.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-18
- Publication Date
- 2025-04-11
AI Technical Summary
Traditional methods are difficult to achieve unified management of multiple equipment resources, are inefficient, and are difficult to achieve automated and refined management.
The ubiquitous resource convergence management method across the cloud-end is adopted to collect device status and energy consumption data in real time through sensors, perform data preprocessing and analysis, establish a fault prediction model, energy consumption optimization model and scheduling optimization model, and conduct model training and optimization through the cloud computing layer and terminal control layer.
It realizes efficient management of equipment resources, significantly improves management efficiency, and can maintain efficient resource scheduling and management in a large number of equipment and complex environments, and realizes automated and refined management.
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of resource fusion management, and in particular to a ubiquitous resource fusion management method and system. Background Art
[0002] The management of smart orchards requires receiving a large amount of equipment information and completing equipment scheduling. Traditional methods are difficult to cope with the unified management of multiple equipment resources, are inefficient, and difficult to achieve automated and refined management. Therefore, the present invention proposes a cross-cloud-end ubiquitous resource fusion management method, combined with artificial intelligence methods such as machine learning, through the powerful computing power of the cloud computing layer and the real-time response of the end control layer, to solve the problem of difficult equipment resource management.
[0003] Chinese invention patent CN102932385 "A business processing method and device for ubiquitous terminals" integrates multiple sensors and communication modules on ubiquitous terminals to collect user and environmental data in real time, and transmits the data to the background server through a wireless network for processing and analysis. The device can automatically adjust the working mode of the terminal according to the analysis results, provide personalized services, and improve user experience. However, since this method relies on a large number of sensors and network communication modules, its implementation cost is high and it is highly dependent on the network environment. In an environment with poor network signals, real-time data transmission and processing may not be achieved, affecting the service quality. Summary of the invention
[0004] In view of the deficiencies in the prior art, the present invention aims to provide a ubiquitous resource fusion management method and system that can achieve automated and refined management.
[0005] To achieve the above object, the present invention can be implemented through the following technical solutions:
[0006] A ubiquitous resource fusion management method, comprising:
[0007] Collect equipment status data and equipment energy consumption data in real time through sensors;
[0008] Pre-process the collected equipment status data and equipment energy consumption data;
[0009] Perform data analysis on the pre-processed data;
[0010] Establish equipment failure prediction models, energy consumption optimization models, and scheduling optimization models based on the data obtained from data analysis;
[0011] According to the data obtained from the data analysis, the fault prediction model, the energy consumption optimization model and the scheduling optimization model are respectively trained and optimized;
[0012] The equipment is responded to and controlled according to the fault prediction model, energy consumption optimization model and scheduling optimization model that have been trained and optimized.
[0013] Furthermore, the preprocessing of the collected device status data and device energy consumption data includes:
[0014] Clean the collected equipment status data and equipment energy consumption data, and use data cleaning tools to remove duplicates, fill missing values, and detect and correct outliers;
[0015] Transform the cleaned data and use ETL tools to convert, standardize and normalize the data.
[0016] Furthermore, the data analysis of the pre-processed data includes:
[0017] Extract key features from the transformed data using feature engineering technology, wherein the key features include time series features, frequency domain features, and statistical features of the equipment operation status;
[0018] The most representative key features are selected through feature selection algorithm;
[0019] The most representative key features are screened out and divided into a training set and a test set. The training set is used for model training, and the test set is used for evaluating model performance.
[0020] Furthermore, the data analysis of the pre-processed data also includes:
[0021] Data enhancement technology is used to deal with the problem of imbalanced training set data.
[0022] Furthermore, the data analysis of the pre-processed data also includes:
[0023] Using data visualization tools and programming languages, the most representative key features are screened out for visual display.
[0024] Furthermore, the training and optimization of the fault prediction model based on the data obtained from the data analysis includes:
[0025] The k-fold cross-validation technique is applied to the training set, and the fault prediction model is established by repeatedly training on different data subsets;
[0026] Verify the trained fault prediction model through the test set to evaluate the stability and performance of the fault prediction model;
[0027] Using grid search and random search methods, the hyperparameters of the fault prediction model are systematically adjusted to find the optimal parameter combination, and the optimal parameter combination is returned to the fault prediction model to improve the prediction accuracy of the fault prediction model.
[0028] Furthermore, the training and optimization of the energy consumption optimization model according to the data obtained from the data analysis includes:
[0029] The k-fold cross-validation technique is applied to the training set, and the energy consumption optimization model is established by repeatedly training on different data subsets;
[0030] The energy consumption optimization model is verified through the test set, the fitting effect of the energy consumption optimization model is evaluated, and the parameters of the energy consumption optimization model are adjusted to ensure the accuracy and stability of energy consumption prediction;
[0031] Use particle swarm optimization and genetic algorithms to optimize the energy consumption optimization model, search for the optimal parameter configuration of equipment operation, and return the optimal parameter configuration to the energy consumption optimization model to reduce energy consumption and improve energy utilization efficiency;
[0032] During the optimization process, safety and performance constraints on equipment operation are set to ensure that energy is saved without affecting the normal operation of the equipment.
[0033] Furthermore, the performing model training and optimization on the scheduling optimization model according to the data obtained from the data analysis includes:
[0034] The k-fold cross-validation technique is applied to the training set, and the scheduling optimization model is established by repeatedly training on different data subsets;
[0035] Use the test set to comprehensively evaluate the scheduling optimization model and adjust the scheduling optimization model parameters to ensure the effectiveness and practicality of the scheduling optimization model;
[0036] Based on the effect of the scheduling optimization model on the test set, adjust the scheduling optimization model parameters to ensure the effectiveness of the scheduling optimization model.
[0037] A ubiquitous resource fusion management system, comprising:
[0038] The terminal control layer interacts with the device to collect device status data and device energy consumption data, and controls the device in real time according to the instructions of the cloud computing layer;
[0039] The edge computing layer is used to pre-process the collected data, and the data pre-processing includes filtering, denoising and format conversion of the collected data;
[0040] The cloud computing layer is used for data storage, data analysis, model building and training optimization of pre-processed data, and transmits the trained and optimized model to the terminal control layer.
[0041] Furthermore, the terminal control layer is also used for:
[0042] Local storage: temporarily storing data when the network is interrupted to ensure data integrity and continuity;
[0043] Security management, including identity authentication, permission control and data encryption, prevents unauthorized access and data leakage.
[0044] Compared with the prior art, the present invention has the following beneficial effects:
[0045] The present invention uses a cross-cloud-end architecture design to fully utilize the powerful computing power of the cloud computing layer and the real-time response capability of the end control layer to achieve efficient management of device resources. Compared with traditional methods, this method can better cope with the unified management of multiple device resources and significantly improve management efficiency, especially when facing a large number of devices and complex environments, it can still maintain efficient resource scheduling and management. DETAILED DESCRIPTION
[0046] The present invention will be further described below in conjunction with specific implementation methods. The terms such as "upper", "inner", "middle", "left", "right" and "one" cited in this specification are only for the convenience of description, and are not used to limit the scope of the present invention. The change or adjustment of their relative relationship shall also be regarded as the scope of the present invention without substantially changing the technical content.
[0047] The present invention is used to manage and optimize the collection, analysis and equipment scheduling of large amounts of environmental data and equipment information, and can be applied to scenarios such as smart orchard management.
[0048] A ubiquitous resource integration management method comprises the following steps:
[0049] S1. Collect equipment status data and equipment energy consumption data in real time through sensors to provide real-time data support for equipment monitoring and scheduling optimization.
[0050] S2. Preprocessing the collected equipment status data and equipment energy consumption data;
[0051] The preprocessing steps include:
[0052] Data cleaning: Clean the collected equipment status data and equipment energy consumption data, and use data cleaning tools (such as Apache Nifi, Talend) to remove duplicates, fill missing values, detect and correct outliers to ensure the accuracy and completeness of the data;
[0053] Data transformation: Transform the cleaned data and use ETL (Extract, Transform, Load) tools (such as Apache Spark and Apache Flink) to convert, standardize and normalize the data to prepare for subsequent analysis.
[0054] S3, performing data analysis on the pre-processed data;
[0055] The steps of data analysis include:
[0056] Use feature engineering technology to extract key features from the transformed data, and use time series analysis, signal processing, spectrum analysis and other methods to extract the timing characteristics, frequency domain characteristics and statistical characteristics of the equipment operation status;
[0057] The most representative key features are selected through feature selection algorithms (such as principal component analysis and LASSO regression), which reduces the data dimension and improves the training effect of the model.
[0058] The most representative key features are screened out and divided into a training set and a test set. The training set is used for model training, and the test set is used to evaluate model performance and ensure the generalization ability of the model.
[0059] Data enhancement techniques (such as oversampling and undersampling) are used to deal with the problem of imbalanced training set data. In particular, when fault data is scarce, oversampling is performed using techniques such as SMOTE to increase the number of fault samples and improve the prediction ability and stability of the model.
[0060] Using data visualization tools (such as Tableau and PowerBI) and programming languages (such as Python's Matplotlib and Seaborn), the most representative key features are screened out for visualization, helping analysts to intuitively understand data characteristics and trends.
[0061] S4. Establish equipment failure prediction model, energy consumption optimization model and scheduling optimization model based on the data obtained from data analysis;
[0062] According to the key features extracted from data analysis, random forest algorithm and support vector machine (SVM) were selected to establish fault prediction model, energy consumption optimization model and scheduling optimization model.
[0063] Random forest algorithm: Select the random forest algorithm, which builds multiple decision trees for ensemble learning and randomly selects features for decision making. It has high accuracy and anti-overfitting capabilities and is suitable for diverse equipment failure data.
[0064] Support Vector Machine (SVM): For linearly inseparable data, support vector machine is selected to handle complex equipment failure modes by constructing the optimal hyperplane in high-dimensional space to maximize the classification boundary.
[0065] S5. Perform model training and optimization on the fault prediction model, energy consumption optimization model and scheduling optimization model respectively according to the data obtained from the data analysis.
[0066] The training and optimization of the fault prediction model includes the following steps:
[0067] Model training: The k-fold cross-validation technique is applied to the training set, and the fault prediction model is established by repeated training on different data subsets;
[0068] Model verification: Verify the trained fault prediction model through the test set to evaluate the stability and performance of the fault prediction model;
[0069] Hyperparameter optimization: Use grid search and random search methods to systematically adjust the hyperparameters of the fault prediction model (such as the number of trees in the random forest and the kernel function of the SVM) to find the optimal parameter combination, and return the optimal parameter combination to the fault prediction model to improve the prediction accuracy of the model;
[0070] Model Evaluation:
[0071] Performance evaluation indicators: Use indicators such as accuracy, precision, recall, and F1 score to comprehensively evaluate the performance of the fault prediction model to ensure the robustness and reliability of the fault prediction model for different fault types;
[0072] Confusion matrix and ROC curve: Use the confusion matrix to analyze the classification results of the model and identify the prediction deviation of the fault prediction model on different fault types. Use the ROC curve to evaluate the binary classification performance of the model and select the optimal decision threshold.
[0073] The training and optimization of the energy consumption optimization model includes the following steps:
[0074] Model training: The k-fold cross-validation technique is applied to the training set, and the energy consumption optimization model is established by repeated training on different data subsets;
[0075] Model verification: verify the energy consumption optimization model through the test set, evaluate the fitting effect of the energy consumption optimization model, adjust the parameters of the energy consumption optimization model, and ensure the accuracy and stability of energy consumption prediction;
[0076] Model optimization: Use particle swarm optimization, genetic algorithm and other optimization algorithms to optimize the energy consumption optimization model, search for the optimal parameter configuration of equipment operation, and return the optimal parameter configuration to the energy consumption optimization model to reduce energy consumption and improve energy utilization efficiency;
[0077] Constraint processing: During the optimization process, set the safety and performance constraints of equipment operation to ensure that energy is saved without affecting the normal operation of the equipment;
[0078] Energy consumption forecast:
[0079] Time series forecasting: Use time series forecasting methods (such as ARIMA model, LSTM neural network) to make short-term and long-term forecasts of the energy consumption of equipment under different operating conditions and capture energy consumption trends and fluctuations;
[0080] Model comparison: By comparing the prediction effects of different time series models, the most suitable model is selected for energy consumption prediction to ensure the accuracy of the prediction results;
[0081] Strategy Implementation:
[0082] Optimization strategy application: Apply the optimized energy consumption strategy to the actual operation of the equipment, monitor the equipment status in real time through the automation system, and adjust the equipment parameters based on the feedback to ensure the effective implementation of the energy consumption optimization strategy;
[0083] Continuous optimization: Based on the energy consumption data in actual operation, the energy consumption optimization model is adjusted regularly and the strategy is updated to adapt to changes in equipment status and external environment.
[0084] The training and optimization of the scheduling optimization model includes the following steps:
[0085] Model training: The k-fold cross-validation technique is applied to the training set, and the scheduling optimization model is established by repeated training on different data subsets;
[0086] Model evaluation: Use the test set to comprehensively evaluate the scheduling optimization model, adjust the model parameters, and ensure the effectiveness and practicality of the scheduling optimization model;
[0087] Model optimization: Based on the effect of the scheduling optimization model on the test set, adjust the scheduling optimization model parameters to ensure the effectiveness of the scheduling optimization model; first, by analyzing the prediction results on the test set, identify the performance differences of the model under different parameter combinations; then, optimize the parameter combination through grid search or random search and other methods to find the optimal model configuration; these parameters may include the number of trees, tree depth, maximum number of features, sample weights, etc.; through continuous iterative adjustment, ensure that the model achieves optimal performance on the test set.
[0088] S6. Responding and controlling the equipment according to the trained and optimized fault prediction model, energy consumption optimization model and scheduling optimization model.
[0089] The present invention combines a fault prediction model, an energy consumption optimization model and a scheduling optimization model to achieve real-time response and control of equipment and ensure efficient operation of the equipment.
[0090] Real-time data collection and preprocessing: Through the sensor network, equipment data is collected and preprocessed in real time to ensure the real-time and accuracy of the data.
[0091] Real-time fault prediction and response: Predict equipment failures in real time and trigger early warnings or automatic repairs when the probability of failure reaches a threshold.
[0092] Real-time energy consumption optimization and control: Adjust equipment operating parameters according to the energy consumption optimization model to achieve the best configuration of energy consumption.
[0093] Real-time scheduling optimization and execution: Dynamically adjust equipment scheduling plans, optimize task priorities and equipment resource allocation, and improve overall efficiency.
[0094] Real-time feedback and iterative optimization: Continuously optimize model parameters through feedback mechanisms to ensure long-term stable operation of the system and achieve decision optimization through human-machine collaboration.
[0095] The ubiquitous resource fusion management system of the present invention mainly includes a cloud computing layer, an edge computing layer, and a terminal control layer.
[0096] 1. Cloud computing layer
[0097] The cloud computing layer is the core of the system, providing computing power and big data processing. It is used for data storage, data analysis, model building and training optimization of pre-processed data, and transmitting the trained and optimized model to the terminal control layer. Its functions include:
[0098] Data storage: Distributed database technology is used to efficiently store pre-processed data, operation logs, and historical records, ensuring high availability and reliability of data, and providing a basis for data analysis and model training.
[0099] Data analysis: Use big data technology to conduct a comprehensive analysis of pre-processed data, extract the key characteristics and operating rules of the equipment, support the training of machine failure prediction models, energy consumption optimization models and scheduling optimization models, and optimize equipment status prediction and scheduling.
[0100] Model establishment: Based on the key features extracted from data analysis, random forest algorithm and support vector machine (SVM) are selected to establish fault prediction model, energy consumption optimization model and scheduling optimization model.
[0101] Model training: The k-fold cross-validation technology is applied to the training set. The fault prediction model, energy consumption optimization model and scheduling optimization model are established by repeated training on different data subsets to improve the intelligent level of equipment management.
[0102] Model optimization: By analyzing the prediction results on the test set, we can identify the performance differences of the model under different parameter combinations; then, we can optimize the parameter combination through grid search or random search to find the optimal model configuration; these parameters may include the number of trees, tree depth, maximum number of features, sample weights, etc.; through continuous iterative adjustments, we can ensure that the model achieves optimal performance on the test set. At the same time, we can verify the generalization ability of the model through cross-validation to ensure that the model is stable and effective in actual application scenarios.
[0103] 2. Edge computing layer
[0104] It is used to pre-process the collected data, and the data pre-processing includes filtering, denoising and format conversion of the collected data. Its functions include:
[0105] Data preprocessing: Filter, denoise and convert the format of the collected data to ensure data quality and lay the foundation for subsequent analysis.
[0106] Data transmission: Data is transmitted to the cloud computing layer through efficient protocols, and data compression and encryption are performed at the same time to ensure the efficiency and security of transmission.
[0107] 3. Terminal control layer
[0108] The terminal control layer interacts with the device to collect device status data and device energy consumption data, and controls the device in real time according to the instructions of the cloud computing layer. Its functions include:
[0109] Device control: Execute the instructions of the cloud computing layer (trained and optimized models) to ensure that the equipment operates according to the optimization strategy to improve efficiency and reliability.
[0110] Data collection: Sensors are used to collect equipment status data and equipment energy consumption data in real time, providing real-time data support for equipment monitoring and scheduling optimization.
[0111] Local storage: Temporarily store data when the network is interrupted to ensure data integrity and continuity.
[0112] Security management: including identity authentication, permission control and data encryption to prevent unauthorized access and data leakage.
[0113] Data transmission and storage
[0114] 1. Efficient transmission
[0115] The system uses efficient data transmission protocols (such as MQTT and HTTP / 2) to ensure that data is quickly transmitted from terminal devices to the cloud computing platform. It uses data compression (such as Gzip), TLS encryption and transmission optimization technology to ensure the efficiency and security of data transmission.
[0116] 2. Data Storage
[0117] On the cloud computing platform, the system uses distributed databases (such as HBase and Cassandra) to efficiently store device data and ensure high data availability. Through data partitioning, replication, compression and archiving, the system improves storage efficiency and fault tolerance.
[0118] Application scenarios of the present invention
[0119] Equipment status monitoring: Real-time monitoring of the operating status of various equipment in the orchard, timely detection and resolution of equipment failures.
[0120] Equipment energy consumption management: Reduce equipment energy consumption and improve energy utilization efficiency through energy consumption optimization model.
[0121] Equipment scheduling optimization: Utilize the scheduling optimization model to achieve efficient use of equipment resources and reduce resource waste.
[0122] Fault prediction and maintenance: Through the fault prediction model, possible equipment failures can be predicted in advance, preventive maintenance can be performed, and equipment downtime can be reduced.
[0123] The present invention has the following advantages:
[0124] Improved resource management efficiency: The present invention uses a cross-cloud-end architecture design to fully utilize the powerful computing power of the cloud computing layer and the real-time response capability of the terminal control layer to achieve efficient management of device resources. Compared with traditional methods, this method can better cope with the unified management of multiple device resources and significantly improve management efficiency, especially when facing a large number of devices and complex environments, it can still maintain efficient resource scheduling and management.
[0125] Automation and refined management: Combining machine learning and big data technology, the equipment failure prediction model, energy consumption optimization model and scheduling optimization model are established, trained and optimized. The invention realizes accurate prediction of equipment status and scheduling optimization, with a high degree of automation and the ability to dynamically adjust strategies according to the equipment operating status. This feature makes smart orchard management more refined, reduces the need for human intervention, and improves the overall management level.
[0126] Real-time response and adaptive optimization: The present invention can quickly respond to changes in device status through real-time analysis and feedback mechanisms at the edge computing and terminal control layers, and continuously improve strategies for device failure prediction, energy consumption optimization, and scheduling optimization through adaptive optimization mechanisms to ensure long-term stable operation of the system. This real-time and adaptive capability has not been effectively addressed in the prior art.
[0127] Reduced dependence and enhanced robustness: Compared with the existing technology, the present invention reduces the dependence on a single hardware device or network environment. Through distributed computing and data processing mechanisms, even in the case of poor network environment or equipment problems, the system can still effectively collect, process and manage data, thereby improving the overall robustness and reliability of the system.
[0128] Cost-effectiveness: Although the present invention introduces technologies such as cloud computing, edge computing and machine learning, its architectural design takes into account the optimal configuration and utilization of resources, reduces dependence on high-cost hardware, optimizes equipment scheduling and energy consumption management, thereby reducing overall operating costs. Compared with some existing technologies that rely on a large number of sensors and communication modules, it has better cost-effectiveness.
[0129] The implementation methods of the present invention are not limited to this. According to the above content of the present invention, using common technical knowledge and customary means in the field, without departing from the above basic technical ideas of the present invention, the present invention can also make other various forms of modification, replacement or combination, all of which fall within the scope of protection of the present invention.
Claims
1. A ubiquitous resource fusion management method, characterized in that: include: Collect equipment status data and equipment energy consumption data in real time through sensors; Pre-process the collected equipment status data and equipment energy consumption data; Perform data analysis on the pre-processed data; Establish equipment failure prediction models, energy consumption optimization models, and scheduling optimization models based on the data obtained from data analysis; According to the data obtained from the data analysis, the fault prediction model, the energy consumption optimization model and the scheduling optimization model are respectively trained and optimized; The equipment is responded to and controlled according to the fault prediction model, energy consumption optimization model and scheduling optimization model that have been trained and optimized.
2. The ubiquitous resource fusion management method according to claim 1, characterized in that: The preprocessing of the collected device status data and device energy consumption data includes: Clean the collected equipment status data and equipment energy consumption data, and use data cleaning tools to remove duplicates, fill missing values, and detect and correct outliers; Transform the cleaned data and use ETL tools to convert, standardize and normalize the data.
3. The ubiquitous resource fusion management method according to claim 2, characterized in that: The data analysis of the pre-processed data includes: Extract key features from the transformed data using feature engineering technology, wherein the key features include time series features, frequency domain features, and statistical features of the equipment operation status; The most representative key features are selected through feature selection algorithm; The most representative key features are screened out and divided into a training set and a test set. The training set is used for model training, and the test set is used for evaluating model performance.
4. The ubiquitous resource integration management method according to claim 3 is characterized in that: The data analysis of the pre-processed data also includes: Data enhancement technology is used to deal with the problem of imbalanced training set data.
5. The ubiquitous resource integration management method according to claim 3 is characterized in that: The data analysis of the pre-processed data also includes: Using data visualization tools and programming languages, the most representative key features are screened out for visual display.
6. The ubiquitous resource integration management method according to claim 3, characterized in that: The training and optimization of the fault prediction model according to the data obtained by data analysis includes: The k-fold cross-validation technique is applied to the training set, and the fault prediction model is established by repeatedly training on different data subsets; Verify the trained fault prediction model through the test set to evaluate the stability and performance of the fault prediction model; Using grid search and random search methods, the hyperparameters of the fault prediction model are systematically adjusted to find the optimal parameter combination, and the optimal parameter combination is returned to the fault prediction model to improve the prediction accuracy of the fault prediction model.
7. The ubiquitous resource integration management method according to claim 3, characterized in that: The training and optimization of the energy consumption optimization model according to the data obtained by data analysis includes: The k-fold cross-validation technique is applied to the training set, and the energy consumption optimization model is established by repeatedly training on different data subsets; The energy consumption optimization model is verified through the test set, the fitting effect of the energy consumption optimization model is evaluated, and the parameters of the energy consumption optimization model are adjusted to ensure the accuracy and stability of energy consumption prediction; Use particle swarm optimization and genetic algorithms to optimize the energy consumption optimization model, search for the optimal parameter configuration of equipment operation, and return the optimal parameter configuration to the energy consumption optimization model to reduce energy consumption and improve energy utilization efficiency; During the optimization process, safety and performance constraints on equipment operation are set to ensure that energy is saved while not affecting the normal operation of the equipment.
8. The ubiquitous resource integration management method according to claim 1, characterized in that: The performing model training and optimization on the scheduling optimization model according to the data obtained from the data analysis comprises: The k-fold cross-validation technique is applied to the training set, and the scheduling optimization model is established by repeatedly training on different data subsets; Use the test set to comprehensively evaluate the scheduling optimization model and adjust the scheduling optimization model parameters to ensure the effectiveness and practicality of the scheduling optimization model; Based on the effect of the scheduling optimization model on the test set, adjust the scheduling optimization model parameters to ensure the effectiveness of the scheduling optimization model.
9. A ubiquitous resource fusion management system, characterized in that: include: The terminal control layer interacts with the device to collect device status data and device energy consumption data, and controls the device in real time according to the instructions of the cloud computing layer; The edge computing layer is used to pre-process the collected data, and the data pre-processing includes filtering, denoising and format conversion of the collected data; The cloud computing layer is used for data storage, data analysis, model building and training optimization of pre-processed data, and transmits the trained and optimized model to the terminal control layer.
10. The ubiquitous resource integration management system according to claim 9, characterized in that: The terminal control layer is also used for: Local storage: temporarily storing data when the network is interrupted to ensure data integrity and continuity; Security management, including identity authentication, permission control and data encryption, prevents unauthorized access and data leakage.