Electrical automatic control system and control method

Abstract

The invention relates to the technical field of electrical automation, in particular to an electrical automatic control system and control method.The electrical automatic control system comprises a hardware part and a software part, and the hardware part comprises an intelligent control module (1), a data preprocessing unit used for preprocessing collected original data; (2) a deep learning network construction unit for designing and constructing a deep learning network architecture suitable for a specific control task; (3) a training and optimizing unit is used for minimizing control errors and improving control precision; (4) a self-learning and evolution unit which continuously adjusts and optimizes a control strategy according to real-time operation data and feedback; the parameter adjusting and strategy updating unit automatically adjusts control parameters and strategies, when facing complex and changeable network environments or service requirements, the system can conduct resource scheduling and management more flexibly and efficiently so as to overcome the defects that in the prior art, fixed control logic and manual intervention are relied on, and the system can not conduct resource scheduling and management more flexibly and efficiently. And complex and changeable working conditions and continuously changing control requirements are difficult to deal with.

Description

Technical Field

[0001] The present invention relates to the field of electrical automation technology, and in particular to an electrical automatic control system and a control method. Background Art

[0002] In the prior art, the control and management of electrical automatic control systems mainly rely on traditional control methods and management tools. Although these methods and tools can help managers to achieve basic control and monitoring of the system to a certain extent, they have many limitations.

[0003] Existing technologies usually use fixed control logic and manual intervention to achieve resource scheduling and management of electrical automatic control systems. This approach is not only inefficient, but also difficult to cope with complex and changing working conditions and changing control requirements. Once the control logic is set, it is often difficult to adjust flexibly and cannot adapt to control requirements in different scenarios.

[0004] Existing technologies also have obvious deficiencies in data collection and processing. Traditional sensor systems can usually only collect limited status information, and the data processing method is relatively simple, lacking in-depth feature extraction and analysis capabilities. This makes it difficult for the system to fully and real-time reflect the operating status and performance of electrical equipment, affecting the accuracy and reliability of control.

[0005] Existing technologies also have shortcomings in anomaly detection and fault handling. Traditional fault detection methods are often based on fixed thresholds or rules, and only when the system status reaches or exceeds these thresholds will an alarm be triggered or corresponding measures be taken. However, this approach often fails to capture potential problems and abnormal conditions in the system in a timely manner, resulting in insufficient timeliness and effectiveness of fault handling.

[0006] In addition, the user interface and interaction mode of the prior art are relatively simple. Usually, only a fixed interface and limited interactive functions can be provided, which cannot meet the personalized control needs and operating habits of users. This may make users feel inconvenient when using the system, reducing the ease of use and user satisfaction of the system.

[0007] In summary, the existing technology has many deficiencies in electrical automatic control systems and control, and cannot meet the current complex and changeable working conditions and ever-changing control requirements. Therefore, a more efficient and intelligent control method is needed to overcome these limitations and improve the control accuracy, efficiency and reliability of the system.

[0008] The control logic of existing electrical automatic control systems is often pre-set. Once the operating conditions change or the control requirements are adjusted, manual intervention is required to modify the control logic, which greatly reduces the flexibility and adaptability of the system. Traditional sensor systems can only collect limited status information, and the data processing method is simple and lacks deep feature extraction and analysis capabilities, which makes it difficult for the system to fully and real-time reflect the operating status of electrical equipment. Existing technologies usually rely on fixed thresholds or rules for anomaly detection and fault handling. This method often cannot capture potential problems and abnormal conditions in the system in a timely manner, affecting the stability and reliability of the system. Summary of the invention

[0009] Technical issues solved In view of the above-mentioned shortcomings of the prior art, the present invention provides an electric automatic control system and a control method, which can effectively solve the problems in the prior art.

[0010] Technical Solution To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention provides an electric automatic control system, comprising a hardware part and a software part, wherein the hardware part comprises: Intelligent control module: ① Data preprocessing unit, used to preprocess the collected raw data; ② Deep learning network construction unit, designed and constructed deep learning network architecture suitable for specific control tasks; ③ Training and optimization unit, minimized control error and improved control accuracy; ④ Self-learning and evolution unit, continuously adjusted and optimized control strategy according to real-time operation data and feedback; ⑤ Parameter adjustment and strategy update unit, automatically adjusted control parameters and strategies; Multimodal sensor fusion module: It includes: ① sensor data acquisition unit, which integrates multiple types of sensors; ② data synchronization and calibration unit; ③ data fusion processing unit, which conducts comprehensive analysis of data from different sources; ④ working condition identification and classification unit; User customized interface module: ① interface element library unit; ② layout editor unit; ③ interactive design tool unit, ④ graphical programming environment unit, providing graphical programming tools and controls; ⑤ user authority management unit; Fault prediction and self-repair module: ① status monitoring unit, real-time monitoring of the system's operating status and parameters; ② fault feature extraction unit; ③ fault prediction model unit, ④ warning signal issuing unit; ⑤ self-repair strategy library unit; ⑥ self-repair execution unit.

[0011] The software part includes: Data collection and processing procedures: ① Data collection by timed interruption; ② Data filtering and processing; ③ Data calibration through a pre-established calibration model.

[0012] Control algorithm program: ① PID control algorithm is used in some scenarios with high control accuracy requirements and clear system models; ② Fuzzy control algorithm is used under other conditions; Communication program: Developed based on the TCP / IP protocol stack, using the Ethernet communication program of the Socket programming interface to send data packets to the server.

[0013] Furthermore, in the intelligent control module, data preprocessing includes cleaning, denoising, normalization and time series alignment; in the deep learning network construction unit, the deep learning architecture specifically includes the selection of network layers, number of neurons, activation functions, etc. to achieve intelligent representation and learning of control strategies. In the training and optimization unit, historical data is used to train the deep learning network, and the back propagation algorithm and optimizer are used to continuously adjust the network parameters. Self-learning and evolution unit: control parameters and strategies are automatically adjusted through online learning. In the parameter adjustment and strategy update unit, automatic adjustment needs to be based on the results of self-learning and evolution to ensure that the system maintains the optimal control state.

[0014] Furthermore, in the multimodal sensor fusion module, the various types of sensors in ① include temperature sensors, pressure sensors, flow sensors, vibration sensors, etc., to realize multi-dimensional data collection; ② in the data synchronization and calibration unit, the data from different sensors are time synchronized and spatially calibrated to ensure the accuracy and consistency of the data; ③ the data from different sources are analyzed by adopting the data fusion algorithm; and finally, ④ the working condition identification and analysis unit can identify the working condition type and characteristics through the fused data, providing an accurate basis for the formulation of the control strategy.

[0015] Furthermore, in the user customized interface module, ① the interface element library unit can provide a wealth of interface elements, such as buttons, sliders, charts, etc., allowing users to select and configure according to actual needs; ② the layout editor unit provides a drag-and-drop layout editing tool, allowing users to easily adjust the position and size of interface elements to achieve personalized interface layout; ③ the interaction design tool unit provides interaction design tools, such as event triggers, action executors, etc., allowing users to define the interaction logic between interface elements; ④ graphical programming tools and controls include logic gates, timers, counters, etc.; ⑤ the user authority management unit can set different user roles and authority levels to ensure the security of the system and the confidentiality of data.

[0016] Furthermore, in the fault prediction and self-repair modules, ① the operating status and parameters include voltage, current, temperature, vibration, etc.; ② the fault feature extraction unit uses signal processing technology and feature extraction algorithm to extract fault features from monitoring data; ③ the fault prediction model unit can build a fault prediction model based on the machine learning algorithm, analyze and predict the extracted fault features, and determine the time and cause of the potential fault; ④ the early warning signal issuing unit can issue a warning signal in time when a potential fault is predicted, and remind the user to handle it through sound, light or text message; ⑤ the self-repair strategy library unit pre-stores a variety of self-repair strategies, such as adjusting control parameters, switching to backup mode, starting redundant equipment, etc.; ⑥ the self-repair execution unit can automatically select and execute the appropriate self-repair strategy according to the results of the fault prediction model and the self-repair strategy library to ensure the continuous and stable operation of the system.

[0017] Furthermore, in the software data acquisition and processing program, ① the timer module of the controller is used to set an interrupt every 200 milliseconds. In the interrupt service program, the data of each sensor is read in turn; ② the data filtering adopts a sliding average filtering algorithm. For each sensor data, a sliding window with a length of 10 is set. After each new data is collected, it is added to the window, and the earliest data in the window is removed. Then, the average value of the data in the window is calculated as the filtered result; ③ the calibration model is a calibration curve for calibrating the sensor output under different test values ​​before leaving the factory. According to the output value of the sensor, calibration is performed by searching the calibration curve to improve the measurement accuracy.

[0018] Furthermore, in the control algorithm program, the fuzzy algorithm is stored in the fuzzy controller. When the input variables of the fuzzy controller are the deviation between the actual data and the set data and the deviation change rate, the two input variables are fuzzified and divided into multiple fuzzy subsets. Fuzzy control rules are formulated based on actual control experience to obtain the actual control quantity for adjustment in actual use. In the PID algorithm, the control quantity is obtained through calculation of the three links of proportion (P), integration (I) and differentiation (D). The proportional link quickly adjusts the control output according to the deviation size; the integral link is used to eliminate the steady-state error of the system; the differential link can predict the deviation change trend and adjust the control output in advance, so that the system can quickly and stably reach the set data.

[0019] To achieve the above objectives, the present invention also introduces an electrical automatic control method, comprising the following steps: S1: Introduce intelligent control algorithms based on deep learning through the intelligent control module, use historical data to train the deep learning network, automatically learn and optimize the control strategy to adapt to the control requirements under different working conditions; S2: Use multimodal sensor fusion modules to collect system status data, integrate multiple types of sensors to achieve multi-dimensional data collection, and each group of sensor modules collects data according to their own set sampling period. Different sampling periods are determined according to the speed of change of various physical quantities in the actual application scenario. The data collected by the sensor is transmitted to the controller through the corresponding interface, and the data from different sources are comprehensively analyzed through sensor fusion technology; S3: Generate control strategies based on feature-extracted data through a deep learning network, apply the generated control actions to the system, and collect system feedback for subsequent strategy optimization; S4: Use reinforcement learning algorithm to optimize the control strategy network parameters according to system feedback, so that the deep learning network has the ability of self-learning and evolution, and can continuously adjust parameters according to the actual operation effect to improve the control accuracy and efficiency. The controller adjusts and optimizes the control strategy according to the difference between the feedback data and the preset target value, adopts closed-loop control, compares the feedback data with the set value, calculates the new deviation and deviation change rate, inputs it into the control algorithm program again, recalculates the control amount, and continuously adjusts the action of the actuator so that the system gradually approaches the preset target value, forming a closed-loop control, thereby continuously improving the control accuracy and stability of the system; S5: Provides a user-customized interface module, designs a configurable user interface, allows users to customize display content, layout and interaction methods according to actual needs, and provides graphical programming tools to meet personalized control needs; S6: Integrated fault prediction and self-repair module, using machine learning algorithms to monitor and analyze system operation status in real time, predict potential faults and issue early warnings. When a fault is detected, it automatically adjusts the control strategy or switches to backup mode to ensure continuous and stable operation of the system.

[0020] Furthermore, the intelligent control module in S1 specifically includes a data preprocessing unit, a deep learning network construction unit, a training and optimization unit, a self-learning and evolution unit, and a parameter adjustment and strategy update unit, which are used to process the collected raw data, design and build a deep learning network architecture, perform network training and optimization, and adjust and optimize the control strategy according to real-time operation data and feedback.

[0021] Furthermore, the multimodal sensor fusion module in S2 specifically includes a sensor data acquisition unit, a data synchronization and calibration unit, a data fusion processing unit, and a working condition identification and classification unit, which are used to realize multi-dimensional data acquisition, synchronize and calibrate data from different sensors, perform comprehensive data analysis, and identify the type and characteristics of the current working condition.

[0022] Furthermore, the user customized interface module in S5 specifically includes an interface element library unit, a layout editor unit, an interaction design tool unit, a graphical programming environment unit and a user authority management unit, which are used to provide rich interface elements, realize personalized interface layout, define the interaction logic between interface elements, create and modify control logic, and set different user roles and authority levels.

[0023] Furthermore, the fault prediction and self-repair module in S6 specifically includes a status monitoring unit, a fault feature extraction unit, a fault prediction model unit, a warning signal issuing unit, a self-repair strategy library unit and a self-repair execution unit, which are used to monitor the operating status and parameters of the system in real time, extract fault features, build and analyze fault prediction models, issue warning signals, pre-store self-repair strategies, and automatically select and execute appropriate self-repair strategies.

[0024] Beneficial Effects Compared with the known public technology, the technical solution provided by the present invention has the following beneficial effects: In the present invention, by introducing an intelligent control algorithm based on deep learning, the system can automatically learn and optimize the control strategy to adapt to the control requirements under different working conditions. This self-learning and adaptive ability enables the system to more flexibly and efficiently schedule and manage resources when facing complex and changing network environments or business needs, thereby overcoming the problem that the existing technology relies on fixed control logic and manual intervention, and is difficult to cope with complex and changing working conditions and changing control requirements.

[0025] The multimodal sensor fusion module is used to collect system status data, and feature extraction is performed through a deep learning network. This method can more comprehensively and in real time reflect the operating status of electrical equipment, provide an accurate and rich data basis for intelligent control, and solve the problems of limited information collected by sensor systems in existing technologies, simple data processing methods, and lack of deep feature extraction and analysis capabilities. At the same time, both the hardware and software architectures adopt modular design, which is easy to expand and upgrade. Users can easily add new sensors, actuators, or modify control algorithms according to actual needs to meet the diverse needs of different application scenarios.

[0026] The use of reinforcement learning algorithms to monitor and analyze the system's operating status in real time can promptly detect and handle potential problems and abnormal conditions in the system. This anomaly detection and fault handling method based on real-time data greatly improves the stability and reliability of the system, improves the steps of anomaly detection and fault handling based on fixed thresholds, and solves the problem of difficulty in timely capturing potential problems and abnormal conditions in the system.

[0027] A configurable user interface and graphical programming tool are designed to allow users to customize display content, layout and interaction methods according to actual needs. This personalized user interface and interaction method improves the system's ease of use and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 is a flowchart of the hardware system of the present invention; Figure 2 is a flow chart of the software system of the present invention; Figure 3 is a flow chart of the method of the present invention; Figure 4 It is a structural flow chart of the closed-loop control algorithm of step S4 in the present invention. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] The present invention will be further described below in conjunction with the embodiments.

[0032] Embodiment 1: An electrical automatic control system, referring to the attached Figure 1 , the hardware part includes: Intelligent control module: ① Data preprocessing unit, used to preprocess the collected raw data; ② Deep learning network construction unit, designed and constructed deep learning network architecture suitable for specific control tasks; ③ Training and optimization unit, minimized control error and improved control accuracy; ④ Self-learning and evolution unit, continuously adjusted and optimized control strategy according to real-time operation data and feedback; ⑤ Parameter adjustment and strategy update unit, automatically adjusted control parameters and strategies; In the intelligent control module, data preprocessing includes cleaning, denoising, normalization and time series alignment; in the deep learning network construction unit, the deep learning architecture specifically includes the selection of network layers, number of neurons, activation functions, etc. to achieve intelligent representation and learning of control strategies. In the training and optimization unit, historical data is used to train the deep learning network, and the back propagation algorithm and optimizer are used to continuously adjust the network parameters. Self-learning and evolution unit: automatic adjustment of control parameters and strategies through online learning. In the parameter adjustment and strategy update unit, automatic adjustment needs to be based on the results of self-learning and evolution to ensure that the system maintains the optimal control state.

[0033] Multimodal sensor fusion module: It includes: ① sensor data acquisition unit, which integrates multiple types of sensors; ② data synchronization and calibration unit; ③ data fusion processing unit, which conducts comprehensive analysis of data from different sources; ④ working condition identification and classification unit; In the multimodal sensor fusion module, the various types of sensors in ① include temperature sensors, pressure sensors, flow sensors, vibration sensors, etc., to realize multi-dimensional data collection; ② In the data synchronization and calibration unit, the data from different sensors are synchronized in time and calibrated in space to ensure the accuracy and consistency of the data; ③ Data from different sources are analyzed by using data fusion algorithms; and finally, the working condition identification and analysis unit ④ can identify the working condition type and characteristics through the fused data, providing an accurate basis for the formulation of control strategies.

[0034] User customized interface module: ① interface element library unit; ② layout editor unit; ③ interactive design tool unit, ④ graphical programming environment unit, providing graphical programming tools and controls; ⑤ user authority management unit; In the user customized interface module, ① the interface element library unit can provide a variety of interface elements, such as buttons, sliders, charts, etc., allowing users to select and configure according to actual needs; ② the layout editor unit provides a drag-and-drop layout editing tool, allowing users to easily adjust the position and size of interface elements to achieve personalized interface layout; ③ the interaction design tool unit provides interaction design tools, such as event triggers, action executors, etc., allowing users to define the interaction logic between interface elements; ④ the graphical programming tools and controls include logic gates, timers, counters, etc.; ⑤ the user authority management unit can set different user roles and authority levels to ensure the security of the system and the confidentiality of data.

[0035] Fault prediction and self-repair module: ① status monitoring unit, real-time monitoring of the system's operating status and parameters; ② fault feature extraction unit; ③ fault prediction model unit, ④ warning signal issuing unit; ⑤ self-repair strategy library unit; ⑥ self-repair execution unit.

[0036] In the fault prediction and self-repair modules, ① the operating status and parameters include voltage, current, temperature, vibration, etc.; ② the fault feature extraction unit uses signal processing technology and feature extraction algorithm to extract fault features from monitoring data; ③ the fault prediction model unit can build a fault prediction model based on the machine learning algorithm, analyze and predict the extracted fault features, and determine the time and cause of potential faults; ④ the early warning signal issuing unit can issue a warning signal in time when a potential fault is predicted, and remind the user to handle it through sound, light or text message; ⑤ the self-repair strategy library unit pre-stores a variety of self-repair strategies, such as adjusting control parameters, switching to backup mode, starting redundant equipment, etc.; ⑥ the self-repair execution unit can automatically select and execute the appropriate self-repair strategy according to the results of the fault prediction model and the self-repair strategy library to ensure the continuous and stable operation of the system. Embodiment 2: In this system, refer to Figure 2 As shown, it also includes the software part, specifically: Data collection and processing procedures: ① Data collection by timed interruption; ② Data filtering and processing; ③ Data calibration through the pre-established calibration model; In the software data acquisition and processing program, ① the timer module of the controller is used to set an interrupt every 200 milliseconds. In the interrupt service program, the data of each sensor is read in turn; ② the data filtering adopts the sliding average filtering algorithm. For each sensor data, a sliding window with a length of 10 is set. After each new data is collected, it is added to the window, and the earliest data in the window is removed. Then the average value of the data in the window is calculated as the filtered result; ③ the calibration model is a calibration curve calibrated for the sensor output under different test values ​​before leaving the factory. According to the output value of the sensor, calibration is performed by looking up the calibration curve to improve the measurement accuracy.

[0037] Control algorithm program: ① PID control algorithm is used in some scenarios with high control accuracy requirements and clear system models; ② Fuzzy control algorithm is used under other conditions; In the control algorithm program, the fuzzy algorithm is stored in the fuzzy controller. When the input variables of the fuzzy controller are the deviation between the actual data and the set data and the rate of change of the deviation, the two input variables are fuzzified and divided into multiple fuzzy subsets. Fuzzy control rules are formulated based on actual control experience to obtain the actual control quantity for adjustment in actual use. The processed data is input into the control algorithm program, and the control algorithm program performs analysis and calculation according to the preset control strategy. For example, in motor speed control, if the fuzzy control algorithm is used, the motor control quantity that needs to be adjusted is obtained through fuzzy reasoning rules according to the motor speed deviation and the rate of change of the deviation; if the PID control algorithm is used, the control quantity is obtained according to the speed deviation through proportional, integral and differential calculations. In the PID algorithm, the control quantity is obtained through the calculation of the three links of proportional (P), integral (I) and differential (D). The proportional link quickly adjusts the control output according to the deviation size; the integral link is used to eliminate the steady-state error of the system; the differential link can predict the deviation change trend and adjust the control output in advance, so that the system can quickly and stably reach the set data.

[0038] Communication program: developed based on TCP / IP protocol stack, using Ethernet communication program of Socket programming interface to send data packets to the server; Example 3: Reference Figure 3-4 As shown, based on the first and second embodiments, an electrical automatic control method is also introduced, as follows: S1: Through the intelligent control module, an intelligent control algorithm based on deep learning is introduced, and the deep learning network is trained using historical data to automatically learn and optimize the control strategy to adapt to the control requirements under different working conditions; the intelligent control module specifically includes a data preprocessing unit, a deep learning network construction unit, a training and optimization unit, a self-learning and evolution unit, and a parameter adjustment and strategy update unit, which are used to process the collected raw data, design and build a deep learning network architecture, perform network training and optimization, and adjust and optimize the control strategy according to real-time operation data and feedback; by introducing an intelligent control algorithm based on deep learning, the system can automatically learn and optimize the control strategy to adapt to the control requirements under different working conditions. This self-learning and adaptive ability enables the system to more flexibly and efficiently perform resource scheduling and management when facing complex and changing network environments or business needs, so as to overcome the problem that the existing technology relies on fixed control logic and manual intervention, and is difficult to cope with complex and changing working conditions and changing control requirements.

[0039] S2: Use multimodal sensor fusion module to collect system status data, integrate multiple types of sensors to realize multi-dimensional data collection, each group of sensor modules collects data according to their own set sampling period, and different sampling periods are determined according to the speed of change of each physical quantity in the actual application scenario. The data collected by the sensor is transmitted to the controller through the corresponding interface, and the data from different sources are comprehensively analyzed through sensor fusion technology; the multimodal sensor fusion module specifically includes a sensor data acquisition unit, a data synchronization and calibration unit, a data fusion processing unit, and a working condition identification and classification unit, which are used to realize multi-dimensional data collection, synchronize and calibrate data from different sensors, conduct data comprehensive analysis, and identify the type and characteristics of the current working condition; use the multimodal sensor fusion module to collect system status data, and extract features through a deep learning network. This method can more comprehensively and real-timely reflect the operating status of electrical equipment, provide an accurate and rich data basis for intelligent control, and solve the problems of limited information collected by the sensor system in the existing technology, simple data processing methods, and lack of deep feature extraction and analysis capabilities. At the same time, both the hardware and software architectures adopt modular design, which is easy to expand and upgrade. Users can easily add new sensors, actuators or modify control algorithms according to actual needs to meet the diverse needs of different application scenarios.

[0040] S3: Generate control strategies based on feature-extracted data through a deep learning network, apply the generated control actions to the system, and collect system feedback for subsequent strategy optimization; S4: Use reinforcement learning algorithm to optimize the control strategy network parameters according to system feedback, so that the deep learning network has the ability of self-learning and evolution, and can continuously adjust parameters according to the actual operation effect to improve the control accuracy and efficiency. The controller adjusts and optimizes the control strategy according to the difference between the feedback data and the preset target value, adopts closed-loop control, compares the feedback data with the set value, calculates the new deviation and deviation change rate, inputs it into the control algorithm program again, recalculates the control amount, and continuously adjusts the action of the actuator to make the system gradually approach the preset target value, forming a closed-loop control, thereby continuously improving the control accuracy and stability of the system; the reinforcement learning algorithm is used to monitor and analyze the system operation status in real time, which can timely discover and handle potential problems and abnormal conditions in the system. This abnormality detection and fault handling method based on real-time data greatly improves the stability and reliability of the system, improves the steps of abnormality detection and fault handling based on fixed thresholds, and the problem of difficulty in timely capturing potential problems and abnormal conditions in the system.

[0041] S5: Provide a user customized interface module and design a configurable user interface to allow users to customize display content, layout and interaction methods according to actual needs, and provide graphical programming tools to meet personalized control needs; the user customized interface module specifically includes an interface element library unit, a layout editor unit, an interaction design tool unit, a graphical programming environment unit and a user authority management unit, which are used to provide rich interface elements, realize personalized interface layout, define the interaction logic between interface elements, create and modify control logic, and set different user roles and authority levels; a configurable user interface and graphical programming tools are designed to allow users to customize display content, layout and interaction methods according to actual needs. This personalized user interface and interaction method improves the system's ease of use and user satisfaction.

[0042] S6: Integrated fault prediction and self-repair module, using machine learning algorithms to monitor and analyze the system operation status in real time, predict potential faults and issue warnings in advance. When a fault is detected, it automatically adjusts the control strategy or switches to the backup mode to ensure continuous and stable operation of the system. The fault prediction and self-repair module specifically includes a status monitoring unit, a fault feature extraction unit, a fault prediction model unit, a warning signal issuance unit, a self-repair strategy library unit and a self-repair execution unit, which are used to monitor the system's operation status and parameters in real time, extract fault features, build and analyze fault prediction models, issue warning signals, pre-store self-repair strategies, and automatically select and execute appropriate self-repair strategies.

[0043] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. An electrical automatic control system, characterized in that: It includes hardware and software parts, where the hardware part includes: Intelligent control module: Introduces intelligent control algorithms with deep learning and trains the automatic learning network; Multimodal sensor fusion module: uses multi-dimensional sensors to collect data according to fixed sampling periods and transmits it to the controller through the corresponding interface; User customized interface module: Design the user interface based on the provided graphical programming tools to achieve personalized control requirements; Fault prediction and self-repair module: Based on automatic learning networks and algorithms, it conducts real-time monitoring and analysis of the operating status, predicts faults and issues early warnings; The software part includes: Data collection and processing procedures: data are collected by means of timed interrupts, and the data are systematically processed through data filtering algorithms and data calibration models; Control algorithm program: PID control algorithm is used in some scenarios with high control accuracy requirements and clear system models; fuzzy control algorithm is used under other conditions; Communication program: Developed based on the TCP / IP protocol stack, using the Ethernet communication program of the Socket programming interface to send data packets to the server.

2. An electrical automatic control system according to claim 1, characterized in that: In the intelligent control module, it includes: ① data preprocessing including cleaning, denoising, normalization and time series alignment processing; ② deep learning network construction unit, the deep learning architecture specifically includes the method of selecting the number of network layers, intelligent representation and learning, ③ training and optimization unit, using historical data to train the deep learning network, and using the back propagation algorithm and optimizer to continuously adjust the network parameters; ④ self-learning and evolution unit: through online learning, automatically adjust the control parameters and strategies, ⑤ in the parameter adjustment and strategy update unit, automatic adjustment needs to be based on the results of self-learning and evolution to ensure that the system maintains the optimal control state.

3. An electrical automatic control system according to claim 1, characterized in that: The multimodal sensor fusion module includes: ① Various types of sensors: specifically including temperature sensors, pressure sensors, flow sensors and vibration sensors, etc., for multi-dimensional data collection; ② Data synchronization and calibration unit: time synchronization and spatial calibration of data from different sensors to ensure data accuracy and consistency; ③ Data fusion processing unit: data from different sources are analyzed by using data fusion algorithms; ④ Working condition identification and analysis unit: the working condition type and characteristics can be identified through the fused data, providing an accurate basis for the formulation of control strategies.

4. An electrical automatic control system according to claim 1, characterized in that: In the user customized interface module, ① interface element library unit: can provide a variety of interface elements, such as buttons, sliders, charts, etc., allowing users to select and configure according to actual needs; ② layout editor unit: including providing drag-and-drop layout editing tools to adjust the position and size of interface elements to achieve personalized interface layout; ③ interaction design tool unit: provides interaction design tools, such as event triggers, action executors, etc., to define the interaction logic between interface elements; ④ graphical programming tools and controls, including logic gates, timers, counters, etc.; ⑤User authority management unit: set different user roles and authority levels.

5. An electrical automatic control system according to claim 1, characterized in that: The fault prediction and self-repair modules include: ① state detection unit: operating status and parameters include voltage, current, temperature, vibration, etc.; ② fault feature extraction unit: using signal processing technology and feature extraction algorithm to extract fault features from monitoring data; ③ fault prediction model unit: building a fault prediction model based on machine learning algorithm, analyzing and predicting the extracted fault features, and judging the time and cause of potential faults; ④ early warning signal issuing unit: predicting potential faults, issuing early warning signals in time to remind users to deal with them; ⑤ a variety of self-repair strategies are pre-stored in the self-repair strategy library unit; ⑥ self-repair execution unit: automatically selecting and executing appropriate self-repair strategies according to the results of the fault prediction model and the self-repair strategy library.

6. An electrical automatic control system according to claim 1, characterized in that: The software data acquisition and processing program includes: ① Data acquisition by timed interruption: using the timer module of the controller, set an interrupt every 200 milliseconds, and read the data of each sensor in turn in the interrupt service program; ② Data filtering adopts a sliding average filtering algorithm: for each sensor data, set a sliding window with a length of 10. After each new data is collected, it is added to the window, and the earliest data in the window is removed, and then the average value of the data in the window is calculated as the filtered result; ③ Data calibration model: before leaving the factory, the calibration curve of the sensor output under different test values ​​is calibrated. According to the output value of the sensor, calibration is performed by looking up the calibration curve to improve the measurement accuracy.

7. An electrical automatic control system according to claim 1, characterized in that: In the control algorithm program, the fuzzy algorithm is stored in the fuzzy controller. When the input variables of the fuzzy controller are the deviation between the actual data and the set data and the deviation change rate, the two input variables are fuzzified and divided into multiple fuzzy subsets. Fuzzy control rules are formulated based on actual control experience to obtain the actual control quantity for adjustment in actual use. In the PID algorithm, the control quantity is obtained through calculation of the three links of proportion (P), integration (I) and differentiation (D). The proportion link quickly adjusts the control output according to the deviation size; the integral link is used to eliminate the steady-state error of the system; the differentiation link can predict the deviation change trend and adjust the control output in advance, so that the system can quickly and stably reach the set data.

8. An electrical automatic control method, characterized in that: include: S1: Introduce intelligent control algorithms based on deep learning through the intelligent control module, use historical data to train the deep learning network, automatically learn and optimize the control strategy to adapt to the control requirements under different working conditions; S2: Use multimodal sensor fusion modules to collect system status data, integrate multiple types of sensors to achieve multi-dimensional data collection, and each group of sensor modules collects data according to their own set sampling period. Different sampling periods are determined according to the speed of change of various physical quantities in the actual application scenario. The data collected by the sensor is transmitted to the controller through the corresponding interface, and the data from different sources are comprehensively analyzed through sensor fusion technology; S3: Generate control strategies based on feature-extracted data through a deep learning network, apply the generated control actions to the system, and collect system feedback for subsequent strategy optimization; S4: Use reinforcement learning algorithm to optimize the control strategy network parameters according to system feedback, so that the deep learning network has the ability of self-learning and evolution, and can continuously adjust parameters according to the actual operation effect to improve the control accuracy and efficiency. The controller adjusts and optimizes the control strategy according to the difference between the feedback data and the preset target value, adopts closed-loop control, compares the feedback data with the set value, calculates the new deviation and deviation change rate, inputs it into the control algorithm program again, recalculates the control amount, and continuously adjusts the action of the actuator so that the system gradually approaches the preset target value, forming a closed-loop control, thereby continuously improving the control accuracy and stability of the system; S5: Provides a user-customized interface module, designs a configurable user interface, allows users to customize display content, layout and interaction methods according to actual needs, and provides graphical programming tools to meet personalized control needs; S6: Integrated fault prediction and self-repair module, using machine learning algorithms to monitor and analyze system operation status in real time, predict potential faults and issue early warnings. When a fault is detected, it automatically adjusts the control strategy or switches to backup mode to ensure continuous and stable operation of the system.

9. An electrical automatic control method according to claim 8, characterized in that: The intelligent control module in S1 specifically includes a data preprocessing unit, a deep learning network construction unit, a training and optimization unit, a self-learning and evolution unit, and a parameter adjustment and strategy update unit, which are used to process the collected raw data, design and build a deep learning network architecture, perform network training and optimization, and adjust and optimize the control strategy based on real-time operation data and feedback.

10. An electrical automatic control method according to claim 8, characterized in that: The multimodal sensor fusion module in S2 specifically includes a sensor data acquisition unit, a data synchronization and calibration unit, a data fusion processing unit, and a working condition identification and classification unit, which are used to realize multi-dimensional data acquisition, synchronize and calibrate data from different sensors, perform comprehensive data analysis, and identify the type and characteristics of the current working condition.

11. An electrical automatic control method according to claim 8, characterized in that: The user customized interface module in S5 specifically includes an interface element library unit, a layout editor unit, an interaction design tool unit, a graphical programming environment unit, and a user authority management unit, which are used to provide rich interface elements, realize personalized interface layout, define the interaction logic between interface elements, create and modify control logic, and set different user roles and authority levels.

12. An electrical automatic control method according to claim 8, characterized in that: The fault prediction and self-repair module in S6 specifically includes a status monitoring unit, a fault feature extraction unit, a fault prediction model unit, a warning signal issuing unit, a self-repair strategy library unit and a self-repair execution unit, which are used to monitor the operating status and parameters of the system in real time, extract fault features, build and analyze fault prediction models, issue warning signals, pre-store self-repair strategies, and automatically select and execute appropriate self-repair strategies.

Images