Multi-source monitoring and cross-domain cooperation electric control cabinet operating system and management method

Abstract

The invention discloses a multi-source monitoring and cross-domain cooperation electric control cabinet operating system and management method. The multi-source monitoring and cross-domain cooperation electric control cabinet operating system comprises an electric control cabinet, multi-source monitoring equipment, a control management module and a server, the multi-source monitoring equipment is provided with a sensor group; the sensor group is arranged in the electric control cabinet to detect various sensing information so as to monitor the running state of the electric control cabinet in real time; the control management module is connected with the multi-source monitoring equipment so as to realize state monitoring, control adjustment, man-machine interaction and operation management of the electric control cabinet according to the sensing information; and the server wirelessly communicates with the control management module to obtain sensing information for big data analysis, and feeds back the sensing information to the control management module to realize cross-domain cooperation when the state is abnormal. Traditional data communication and modern remote cooperation technologies are combined, monitoring and maintenance of the whole life cycle of the electric control cabinet are achieved, the equipment performance is remarkably improved, the operation cost is reduced, and good application value is achieved.

Description

Technical Field

[0001] The present invention relates to the field of electric control cabinets, and in particular to an electric control cabinet operating system and management method for multi-source monitoring and cross-domain collaboration. Background Art

[0002] An electric control cabinet is an electrical device used for centralized control and monitoring of power equipment. It is usually made of metal materials and is equipped with various electrical components such as transformers, circuit breakers, contactors, relays, and various monitoring and protection devices. Its main functions are to realize power distribution, control operations, ensure the safe operation and stability of equipment, and provide interfaces that are easy to operate and maintain. Electric control cabinets are often used in industrial automation, building electrical and distribution systems. However, with the widespread deployment of electric control cabinets in various applications, the traditional manual inspection method has become increasingly cumbersome, with a huge workload, and it is difficult to detect and handle equipment failures in a timely manner. In addition, the current electric control cabinets have deficiencies in the human-computer interaction function with operators, the diversity of detection parameters, and the real-time monitoring capability, and the remote collaboration function is also not perfect. Therefore, it is urgent to develop an electric control cabinet system with multi-source monitoring and cross-domain collaboration functions that can realize real-time monitoring and fault warning of the operating status of the electric control cabinet to ensure its safe and stable operation. Summary of the invention

[0003] The main purpose of the present invention is to overcome the above-mentioned defects in the prior art, and to propose an electric control cabinet operating system and management method with multi-source monitoring and cross-domain collaboration, which combines traditional data communication and modern remote collaboration technology, realizes the monitoring and maintenance of the electric control cabinet throughout its life cycle, significantly improves equipment performance, and reduces operating costs, and has good application value.

[0004] The present invention adopts the following technical solution:

[0005] A multi-source monitoring and cross-domain collaboration electric control cabinet operating system comprises an electric control cabinet, a multi-source monitoring device, a control management module and a server; the multi-source monitoring device is provided with a sensor group, and the sensor group is arranged in the electric control cabinet to detect a variety of sensor information to realize real-time monitoring of the operating status of the electric control cabinet; the control management module is connected to the multi-source monitoring device to realize status monitoring, control and adjustment, human-computer interaction and operation management of the electric control cabinet according to the sensor information; the server communicates wirelessly with the control management module to obtain the sensor information for big data analysis and feeds back to the control management module when the status is abnormal to realize cross-domain collaboration.

[0006] The sensor group integrates one or more of a temperature sensor, a voltage sensor, a current sensor, a humidity sensor, a vibration sensor, a smoke sensor and a camera.

[0007] The multi-source monitoring device also includes a signal processing module, a CPU module and a communication module; the signal processing module converts the signal data detected by the sensor group into an analog signal, reduces the dimension and normalizes the analog signals of different dimensions, and then amplifies the signal, filters the signal, denoises and smoothes it, and then converts it into a digital signal and inputs it into the CPU module; the control management module is connected to the CPU module for communication; the server exchanges data with the CPU module through the communication module.

[0008] The control management module is provided with an industrial computer and a human-computer interaction device; the industrial computer is connected to the electrical components in the electric control cabinet, and the human-computer interaction module communicates with the industrial computer to realize status monitoring, control adjustment, human-computer interaction and operation management.

[0009] The server adopts the Hadoop distributed computing framework and utilizes the TensorFlow deep learning framework to build multiple neural network models to achieve fault diagnosis, life prediction and performance evaluation.

[0010] A multi-source monitoring and cross-domain collaborative electric control cabinet management method, comprising:

[0011] Real-time detection of various sensor information of the electric control cabinet to monitor the operating status of the electric control cabinet in real time; based on the sensor information, status monitoring, control and adjustment, human-computer interaction and operation management of the electric control cabinet are realized; the sensor information is remotely acquired for big data analysis and fed back to the control management module to realize cross-domain collaboration when the status is abnormal.

[0012] Remotely obtain the sensor information for big data analysis and feed it back to the control management module when the status is abnormal to achieve cross-domain collaboration, specifically including the following:

[0013] Acquire real-time infrared image data of the electric control cabinet, and use a pre-trained convolutional neural network to extract key features from the real-time infrared image data to identify the fault type of the electric control cabinet;

[0014] Acquire real-time sensor information of the electric control cabinet, and input the sensor data into a pre-trained long short-term memory network to predict the remaining service life of the electrical equipment of the electric control cabinet;

[0015] The performance indicators of the electric control cabinet are obtained, and the performance indicators are input into a pre-trained deep neural network to output a comprehensive performance score of the electric control cabinet.

[0016] The convolutional neural network includes multiple convolutional layers, pooling layers and fully connected layers; the low-level features in the infrared image data are extracted through the convolutional layers, the feature dimensions are reduced through the pooling layers, and finally the fault type of the electric control cabinet is identified by classification through the fully connected layers; the cross entropy loss function and Adam optimizer are used during the training of the convolutional neural network.

[0017] The long short-term memory network includes an input layer, multiple LSTM layers and an output layer; the input layer inputs the statistical characteristics of the sensor information and the frequency domain characteristics after Fourier transformation, and the LSTM layer effectively captures the trend of equipment degradation by memorizing long-term dependencies, thereby accurately predicting the remaining service life and outputting it through the output layer; the number of hidden units of the long short-term memory network is set to 128, the mean square error is used as the loss function, and the RMSprop optimizer is used for training.

[0018] The deep neural network includes an input layer, multiple fully connected layers and an output layer, the input layer inputs a performance indicator; each of the fully connected layers uses a ReLU activation function and unifies the scales of different performance indicators through normalization processing; the output layer uses a Softmax activation function to generate a comprehensive performance score of the electric control cabinet; the cross entropy loss function and Adam optimizer are used during deep neural network training, and the learning rate adopts a learning rate decay strategy.

[0019] It can be seen from the above description of the present invention that, compared with the prior art, the present invention has the following beneficial effects:

[0020] 1. In the present invention, multi-source monitoring equipment, control management module and server are used to monitor the operating status of the electric control cabinet in real time through the multi-source monitoring equipment; the control management module cooperates with the multi-source monitoring equipment to realize the real-time operation of the electric control cabinet in combination with the near and far ends, including functions such as status monitoring, control adjustment, human-computer interaction and operation management; the cross-domain collaboration function is realized through the interactive communication between the server and the control management module, including remote control, real-time fault diagnosis, predictive maintenance and system performance optimization, etc., and the traditional data communication and modern remote collaboration technology are combined to realize the monitoring and maintenance of the electric control cabinet throughout its life cycle, which significantly improves the equipment performance and reduces the cost, and has good application value.

[0021] 2. In the present invention, multiple sensors such as temperature sensors, humidity sensors, voltage sensors, vibration sensors and smoke sensors are integrated to comprehensively monitor the working status in the cabinet, ensure comprehensive control of the equipment, timely discover potential faults, and ensure stable operation of the equipment. Human-computer interaction equipment allows operators to access monitoring data more intuitively, perform status detection and control adjustments, and user-friendly interface design improves operational convenience, lowers the threshold for use, and speeds up fault response. At the same time, modern communication means are used in combination with big data analysis technology to achieve remote collaboration of electric control cabinets.

[0022] 3. In the present invention, the server adopts the Hadoop distributed computing framework and uses the TensorFlow deep learning framework to build multiple neural network models to achieve fault diagnosis, life prediction and performance evaluation; it can not only detect possible faults of the electric control cabinet during operation, but also predict the trend of potential problems based on historical operation data and real-time monitoring data; the system also realizes model integration, integrates the prediction results of multiple models, and improves the overall accuracy and robustness.

[0023] 4. The present invention also integrates learning and optimization functions based on big data. The background analysis module will continuously learn the system's operating mode and the user's operating habits, and dynamically adjust the accuracy of the diagnostic model and prediction algorithm to more efficiently capture potential problems and optimize system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the overall structure of the system of the present invention;

[0025] Figure 2 This is a schematic diagram of data communication of the system of the present invention;

[0026] Figure 3 This is a schematic diagram of remote collaborative information transmission according to the present invention.

[0027] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. DETAILED DESCRIPTION

[0028] The present invention is further described below through specific implementation modes.

[0029] See also Figures 1 to 3 , an electric control cabinet operating system with multi-source monitoring and cross-domain collaboration, including an electric control cabinet, a multi-source monitoring device, a control management module and a server. The multi-source detection device is a proximal module, and is provided with a sensor group. The sensor group is arranged in the electric control cabinet to detect a variety of sensor information to realize real-time monitoring of the operating status of the electric control cabinet. The sensor group is the perception nerve of the system, including multiple types of sensors, such as one or more of temperature sensors, voltage sensors, current sensors, humidity sensors, vibration sensors, smoke sensors and cameras. The collected information includes temperature, voltage, current, humidity, vibration signals, smoke signals and images.

[0030] Among them, the temperature sensor can use PT100 platinum resistance. Or the temperature sensor and camera functions can be realized by an infrared camera alone. The infrared camera is installed above the door of the electric control cabinet to monitor the temperature of the components in the cabinet and monitor the operation of the cabinet. The voltage sensor is installed in parallel on the corresponding components that need to strictly monitor the voltage. The current sensor uses a Hall effect sensor. The humidity sensor can be attached to the cabinet wall near the cable connector to monitor the humidity in real time. It can use a capacitive sensor. The vibration sensor uses a MEMS accelerometer and is installed on the load-bearing plate at the bottom of the electric control cabinet. If vibration occurs, the resistor on the vibration sensor will generate a strain signal. The smoke sensor is installed at the top of the electric control cabinet. When the components fail and produce smoke, the sensor can send an alarm signal to the operator in time. These sensors are arranged in key positions of the electric control cabinet, such as around the main control board, power module and cooling system, to monitor the operation status of the electric control cabinet in all directions.

[0031] The raw data collected by the sensor is an analog signal, and the sampling frequency can reach 1kHz to ensure that instantaneous changes are captured. The multi-source monitoring equipment also includes a signal processing module, a CPU module and a communication module. The signal processing module converts the signal data detected by the sensor group into an analog signal, reduces the dimension and normalizes the analog signals of different dimensions, then amplifies the signal, filters the signal, removes noise and smoothes it, and then converts it into a digital signal and inputs it into the CPU module.

[0032] Specifically, the signal processing module of the multi-source monitoring equipment can be implemented using a high-performance digital signal processor (DSP) chip, such as TI's TMS320 series. The module first amplifies the analog signal output by the sensor and uses a precision operational amplifier (such as AD8628) to amplify the weak signal to a suitable level. Subsequently, a bandpass filter is used to remove high-frequency noise and low-frequency drift. The filter adopts a Butterworth design, and the cutoff frequency is optimized according to the characteristics of different sensors. Finally, an adaptive filtering algorithm (such as the minimum mean square error LMS algorithm) is used for denoising to effectively improve the signal-to-noise ratio of the signal. The processed signal is converted into a digital signal through a 16-bit ADC, laying the foundation for subsequent analysis.

[0033] The CPU module, i.e. the central processor, uses Siemens S7-1200 PLC, which is a high-performance programmable logic controller with powerful computing power and rich I / O interfaces. The PLC runs the control program developed by TIAPortal programming software, which uses a combination of ladder diagrams and structured text programming. The program includes multiple functional modules such as data acquisition, status monitoring, and alarm triggering, and achieves rapid response to abnormal situations through interrupt mechanisms. The scan cycle of the PLC is set to 10ms to ensure that the system can process incoming data and control instructions in real time. The communication module can use an RS-485 wireless communication module, specifically the XBee Pro series of MaxStream. The module operates in the 2.4GHz ISM band and uses FHSS (frequency hopping spread spectrum) technology, which has strong anti-interference capabilities. The communication uses the AES-128 encryption algorithm to ensure the security of data transmission. The module supports point-to-point and star network topologies with a maximum transmission distance of up to 15km (under line of sight conditions), meeting the needs of large industrial environments. The communication protocol adopts Modbus RTU, the data frame format is: 1 start bit, 8 data bits, 1 parity bit, 1 stop bit, and the baud rate is adjustable, up to 115200bps.

[0034] The control management module is connected to the CPU module of the multi-source monitoring device to realize the status monitoring, control adjustment, human-computer interaction and operation management of the electric control cabinet according to the sensor information. As a mid-end module, the control management module can directly realize the near-end operation at the electric control cabinet end through data communication, and can also develop corresponding software to realize remote operation at the mobile end. The data communication between the control management module and the multi-source monitoring device includes but is not limited to the controller area network (CAN) and serial communication (RS-485), which has the advantages of efficient communication and low cost. Among them, the control management module is equipped with an industrial computer and a human-computer interaction device. The industrial computer is connected to the electrical components in the electric control cabinet. It can use an Intel Core i5 processor, equipped with 8GB RAM and 256GB SSD storage space, and run the Windows 10 IoT Enterprise operating system. The human-computer interaction module communicates with the industrial computer to realize status monitoring, control adjustment, human-computer interaction and operation management. The human-computer interaction device includes a 15-inch touch screen installed on the inside of the front door of the electric control cabinet. The display screen uses capacitive touch technology with a resolution of 1920×1080. The status monitoring function is realized through the self-developed monitoring software, which is written in C# language and has functions such as real-time data display, historical data query, and trend chart drawing. The control and regulation function is realized through the configured PLC system, using Siemens S7-1200 series PLC, and communicating with the industrial computer through the Profinet protocol. The operation management function includes modules such as equipment management, task scheduling, and alarm management, and uses the SQLite local database to store relevant information.

[0035] The server communicates wirelessly with the control management module to obtain sensor information for big data analysis and feedback to the control management module when the status is abnormal to achieve cross-domain collaboration. As a remote module, the server uses low-latency 5G communication technology to communicate with the communication module, thereby realizing data interaction with the CPU module, which is used to receive big data running near the electric control cabinet. Through the development of corresponding analysis modules in the background, real-time remote collaborative operation of the electric control cabinet can be performed. The server's database uses Microsoft SQL Server 2019 and runs on a high-performance server. The database design adopts a star schema, which includes fact tables and multiple dimension tables, and optimizes query performance. Use stored procedures and triggers to achieve real-time data processing and anomaly detection. The database is also configured with automatic backup and failover mechanisms to ensure data security and high availability of the system.

[0036] The server of the present invention is a cloud server, on which a background analysis system is deployed. The server can be configured with an 8-core CPU, 32GB memory, and 2TB storage space. Data analysis uses the Hadoop distributed computing framework, and historical monitoring data is stored through HDFS. The system uses advanced machine learning technology, is developed based on the Python language, and uses the TensorFlow deep learning framework to build multiple neural network models, mainly including three core functional modules: fault diagnosis, life prediction, and performance evaluation.

[0037] Based on this, the present invention also proposes a multi-source monitoring and cross-domain collaborative electric control cabinet management method, which is implemented by using the above-mentioned multi-source monitoring and cross-domain collaborative electric control cabinet management system, including:

[0038] Real-time detection of various sensor information of the electric control cabinet to monitor the operating status of the electric control cabinet in real time. The sensor information comes from the sensor group of the multi-source monitoring equipment. The information obtained includes infrared image data, etc.; based on the sensor information, the status monitoring, control adjustment, human-computer interaction and operation management of the electric control cabinet are realized; remote acquisition of sensor information for big data analysis and feedback to the control management module when the status is abnormal to realize cross-domain collaboration. For example, if an abnormal status occurs, the remote server will feedback the abnormal status signal to the operator, and the operator can handle the abnormal status at the electric control cabinet end or through the mobile end.

[0039] Among them, remote acquisition of sensor information for big data analysis and feedback to the control management module when the status is abnormal to achieve cross-domain collaboration, specifically including the following:

[0040] Acquire real-time infrared image data of the electric control cabinet, and use a pre-trained convolutional neural network to extract key features from the real-time infrared image data to identify the fault type of the electric control cabinet;

[0041] Obtain real-time sensor information from the electric control cabinet and input the sensor data into a pre-trained long short-term memory network to predict the remaining service life of the electrical equipment in the electric control cabinet;

[0042] The performance indicators of the electric control cabinet are obtained and input into the pre-trained deep neural network to output the comprehensive performance score of the electric control cabinet.

[0043] Among them, fault diagnosis is implemented using a convolutional neural network (CNN), which includes multiple convolutional layers, pooling layers, and fully connected layers. The CNN model receives image data from an infrared camera, which is stored in HDFS. The convolutional layer extracts low-level features in the infrared image data, such as edges and textures. The pooling layer reduces the feature dimension, and finally the fully connected layer performs classification to identify the fault type of the electric control cabinet. The convolutional neural network is trained using a cross entropy loss function and an Adam optimizer, with a learning rate set to 0.001 and a batch size of 64. After training, CNN can extract key features from real-time data, such as temperature anomalies and component wear, to accurately identify the fault type of the electric control cabinet.

[0044] Life prediction is implemented using a long short-term memory network (LSTM), which includes an input layer, multiple LSTM layers, and an output layer. The input layer inputs the statistical features of sensor information and the frequency domain features after Fourier transformation. The statistical features include rolling mean, standard deviation, peak value, and other statistical features. The LSTM layer effectively captures the trend of equipment degradation by memorizing long-term dependencies, thereby accurately predicting the remaining service life and outputting it through the output layer. The number of hidden units in the long short-term memory network is set to 128, and the mean square error is used as the loss function. The RMSprop optimizer is used for training. The LSTM model is trained using the time series data of the device, including sensor data such as temperature, vibration, and current.

[0045] The performance evaluation is implemented using a deep neural network (DNN), which includes an input layer, multiple fully connected layers, and an output layer. The input layer inputs performance indicators, which can be key performance indicators (KPIs) of the electric control cabinet, such as energy efficiency, product quality, equipment utilization, etc. Each fully connected layer uses the ReLU activation function and normalization to unify the scales of different performance indicators, and uses batch normalization technology to improve training stability; the output layer uses the Softmax activation function to generate a comprehensive performance score of the electric control cabinet. The cross entropy loss function and Adam optimizer are used for model training, and the learning rate adopts the learning rate decay strategy.

[0046] The three models of the present invention all adopt an automatic update mechanism, and through online learning technology, the models are regularly retrained using newly collected data. Transfer learning technology is used in the update process to retain the weights of previous training and speed up the convergence speed. At the same time, the system also realizes model integration, integrates the prediction results of multiple models, and improves the overall accuracy and robustness. In terms of data preprocessing, the system uses a variety of technologies, including data cleaning, outlier detection, missing value interpolation, etc. In terms of feature engineering, in addition to traditional statistical features, advanced feature extraction methods such as wavelet transform and principal component analysis (PCA) are also used. In order to deal with data imbalance problems, the system uses oversampling and undersampling techniques, such as the SMOTE algorithm.

[0047] The remote control collaboration software of the present invention adopts a B / S architecture, and the back-end uses the Spring Boot framework to implement a RESTful API and support a microservice architecture. The front-end adopts the Vue.js framework to implement responsive design and component-based development. The mobile APP is implemented through the Flutter cross-platform framework to ensure a consistent experience on iOS and Android systems. The APP adopts the Material Design design language to provide an intuitive user interface and a smooth interactive experience. The inter-system communication adopts the MQTT protocol, which is a lightweight publish-subscribe message transmission protocol that is particularly suitable for the Internet of Things scenario. The MQTTbroker uses Eclipse Mosquitto and is configured with a quality of service (QoS) level of 1 to ensure that the message is delivered at least once.

[0048] In order to handle a large number of concurrent connections, the system also integrates Apache Kafka as a message queue to improve the throughput and reliability of data processing. In terms of security, all data transmission is encrypted with SSL / TLS using the AES-256 encryption algorithm. User access requires two-factor authentication, including passwords and dynamic verification codes. The system also implements role-based access control (RBAC) and refines the permission management of different users. To prevent SQL injection and cross-site scripting (XSS) attacks, the system strictly verifies and filters all user input. In terms of data storage, the system adopts a hybrid storage strategy. Real-time data is stored in the InfluxDB time series database, and historical data is stored in the Hadoop Distributed File System (HDFS). In order to improve query efficiency, the system uses ElasticSearch as a full-text search engine, and cooperates with Logstash and Kibana to build a complete ELK log analysis platform. In order to improve the reliability and scalability of the system, the entire architecture uses containerization technology. Docker is used to containerize various service components, and containers are orchestrated and managed through Kubernetes.

[0049] The system also implements an automated CI / CD process, using Jenkins for continuous integration and deployment to ensure rapid iteration and version control. In terms of visualization, the system integrates Grafana as a data visualization tool, providing a rich set of dashboards and charts that enable users to intuitively monitor device status and performance indicators. At the same time, the system has also developed a 3D visualization module, using the Three.js library to implement 3D model display and interactive operation of the equipment. To support large-scale data analysis, the system integrates the Apache Spark distributed computing framework, which is capable of performing complex batch and stream processing tasks. Using the Spark MLlib machine learning library, the system is able to perform more advanced data mining and predictive analysis tasks, such as anomaly detection, trend analysis, etc.

[0050] The system also provides an open API interface to support integration with other enterprise systems such as ERP, MES, etc., to achieve seamless data flow and automation of business processes. Through this comprehensive and advanced system design, not only intelligent monitoring and predictive maintenance of equipment are achieved, but also strong technical support is provided for the digital transformation of enterprises.

[0051] The system and method of the present invention work as follows:

[0052] The sensor group of the multi-source monitoring equipment collects analog signals and enters the signal processing module for processing. The signal passes through the amplifier circuit (adjustable gain, typical value is 100) and then passes through the 8th order Butterworth bandpass filter (passband is 10Hz-1kHz). The filtered signal is then denoised by the adaptive LMS algorithm, with the algorithm step parameter μ set to 0.01 and the filter order to 32. The processed signal is converted to a digital signal through a 16-bit ADC (such as AD7606), and the sampling rate is set to 10kHz to capture rapidly changing states.

[0053] The digital signal is input to the analog input terminal of the CPU module. A data acquisition function block is set in the PLC program to read all sensor data every 10ms and perform preliminary data verification and standardization.

[0054] The CPU module packages the processed data into Modbus RTU data frames and sends them to the communication module through its built-in RS-485 interface. The data frame contains information such as device ID, timestamp, and readings of each sensor, with a total length of no more than 256 bytes. After receiving the data, the communication module encrypts the data using the AES-128 encryption algorithm. The encryption key is automatically updated every hour to increase security. The module then uses FHSS technology to frequency hop and send data in 79 channels in the 2.4GHz frequency band, with the transmit power set to 18dBm to achieve the maximum transmission distance.

[0055] After receiving the encrypted data, the receiving module on the server side decrypts it and transmits it to the server through the RS-232 interface. The data receiving program (developed in C#) running on the server parses the data and stores it in a temporary table in the SQL Server database.

[0056] The stored procedure in the database is triggered once a second to process the data in the temporary table. The processing includes unit conversion, outlier detection (using the 3σ principle), status judgment, etc. The processed data is stored in the main data table, and the relevant statistical information and status flags are updated at the same time.

[0057] If an abnormal state is detected, the database trigger will immediately generate an alarm message and push the alarm to the application server through the database link service. After receiving the alarm, the application server (developed using ASP.NET Core) pushes it to the client in real time through WebSocket, and sends it to the communication module of the multi-source monitoring device through the MQTT protocol for transmission back to the electric control cabinet.

[0058] After receiving the feedback signal, the communication module of the multi-source monitoring device sends the data back to the communication module at the electric control cabinet end through the same encryption and transmission process. The communication module at the electric control cabinet end receives and decrypts the data and transmits it to the communication processing function block of the CPU module through the RS-485 interface. The CPU module analyzes the feedback signal through the program and performs corresponding control operations according to the preset logic, such as triggering the alarm light, adjusting the equipment parameters, etc. At the same time, the PLC transmits the feedback information to the human-computer interaction device through the Ethernet interface.

[0059] The human-computer interaction module is an industrial-grade touch screen PC running Windows 10 IoT Enterprise operating system. The PC is installed with specially developed monitoring and control software, which is developed in C# language and .NET Framework 4.8 and uses WPF (Windows Presentation Foundation) framework to build the user interface.

[0060] The monitoring and control software on the human-machine interaction module establishes a real-time data connection with the PLC through the OPC UA protocol. The operator can interact with the software through the touch screen or an external keyboard and mouse device. The software supports multi-level user authority management, and users of different levels can access different functional modules. The software also integrates remote assistance functions, which are implemented through the TeamViewer SDK, allowing remote technical support personnel to take over the system for fault diagnosis and remote operation when needed.

[0061] To ensure data integrity and traceability, the system implements a comprehensive logging mechanism. All operational behaviors, state changes, and system events are recorded in the local database and synchronized to the central server regularly. The system also supports offline mode operation. When communication with the central server is interrupted, the local PLC and PC will continue to perform data acquisition and basic control functions, and temporarily store the data locally. After communication is restored, the accumulated data will be automatically synchronized. In order to adapt to different working environments, the human-computer interaction module adopts an industrial-grade design that is dustproof, waterproof, and shockproof, and meets the IP65 protection level. The display uses high-brightness LED backlight technology, which can display clearly even in strong light environments.

[0062] The system of the present invention adopts a highly integrated modular design, covering a top-level receiving module, a background analysis module and remote control collaboration software. The modules collaborate efficiently and are powerful, supporting comprehensive monitoring, fault diagnosis, predictive maintenance and remote control operations of electric control cabinets.

[0063] The receiving module of the server side of the present invention uses the excellent performance Quectel RM500Q 5G module, which has stable and reliable 5G communication capabilities, supports high-speed data transmission, ultra-low latency and big data throughput, and can quickly receive a large amount of real-time data from the remote background database, thereby ensuring the data interaction smoothness and timely response of the electric control cabinet monitoring system. The background analysis module integrates the data collected by multi-source sensors and uses predictive maintenance (PdM) technology to build a powerful intelligent analysis platform. Relying on a large database, the module uses advanced machine learning algorithms to deeply mine and analyze the data. It can not only detect possible faults in the operation of the electric control cabinet, but also predict the trend of potential problems based on historical operation data and real-time monitoring data. For example, when certain key parameters gradually deviate from the normal range, the module will identify risks in advance and generate targeted predictive maintenance suggestions. In addition, the background analysis module can also optimize the system performance of the electric control cabinet, and propose performance optimization strategies by analyzing the overall operating efficiency and load distribution, such as adjusting the fan speed, optimizing the power supply load, etc., to improve the reliability and economy of equipment operation.

[0064] At the data flow and operation level, the backend database achieves low-latency data upload through the 5G network, and transmits the electric control cabinet operation data collected by the sensor group to the server-side receiving module in real time. After receiving this data, the receiving module on the server side quickly downloads and forwards it to the backend analysis module for processing. After the backend analysis is completed, the fault diagnosis information, maintenance suggestions and optimization strategies generated will be sent to the remote control collaboration software synchronously. The software displays the operating status of the electric control cabinet in real time with a user-friendly interface, and provides users with detailed diagnostic reports, including fault location, fault type, impact degree and recommended treatment solutions.

[0065] The predictive maintenance part will remind users to replace key components or arrange maintenance plans in advance based on future risk predictions to prevent unexpected downtime. The remote control collaboration software is not only a platform for information display, but also has built-in communication functions for directly controlling the electric control cabinet. By establishing a secure and stable network connection with the communication module of the electric control cabinet, users can remotely perform a variety of control operations, such as turning off the power of the electric control cabinet, starting or turning off the exhaust fan, and adjusting the parameters of the cooling system. This function enables users to respond quickly in emergencies without on-site operation, greatly improving the convenience of operation and the flexibility of management. At the same time, the software supports multi-terminal collaborative operations. Whether on a computer, mobile phone or tablet device, users can remotely monitor and control the electric control cabinet.

[0066] In order to further improve the reliability and intelligence level of the system, this system also integrates learning and optimization functions based on big data. The background analysis module will continuously learn the system's operating mode and the user's operating habits, and dynamically adjust the accuracy of the diagnostic model and prediction algorithm to more efficiently capture potential problems and optimize system performance. In addition, the system also supports multi-level authority management, and different users can perform corresponding operations according to the authorization scope. For example, ordinary users can only view status information, while advanced users can perform complex remote control or configuration operations. Through these detailed designs, the electric control cabinet operating system of the present invention realizes the organic integration of remote monitoring, intelligent analysis, predictive maintenance and efficient control, and provides an intelligent, networked and efficient solution for industrial equipment management.

[0067] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited thereto. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.

Claims

1. An electric control cabinet operating system for multi-source monitoring and cross-domain collaboration, comprising an electric control cabinet, characterized in that: It also includes a multi-source monitoring device, a control management module and a server; the multi-source monitoring device is provided with a sensor group, and the sensor group is arranged in the electric control cabinet to detect a variety of sensor information to realize real-time monitoring of the operating status of the electric control cabinet; the control management module is connected to the multi-source monitoring device to realize status monitoring, control and adjustment, human-computer interaction and operation management of the electric control cabinet according to the sensor information; the server communicates wirelessly with the control management module to obtain the sensor information for big data analysis and feeds back to the control management module when the status is abnormal to realize cross-domain collaboration.

2. The electric control cabinet operating system for multi-source monitoring and cross-domain collaboration according to claim 1, characterized in that: The sensor group integrates one or more of a temperature sensor, a voltage sensor, a current sensor, a humidity sensor, a vibration sensor, a smoke sensor and a camera.

3. The electric control cabinet operating system for multi-source monitoring and cross-domain collaboration according to claim 1, characterized in that: The multi-source monitoring device also includes a signal processing module, a CPU module and a communication module; the signal processing module converts the signal data detected by the sensor group into an analog signal, reduces the dimension and normalizes the analog signals of different dimensions, and then amplifies the signal, filters the signal, denoises and smoothes it, and then converts it into a digital signal and inputs it into the CPU module; the control management module is connected to the CPU module for communication; the server exchanges data with the CPU module through the communication module.

4. The electric control cabinet operating system for multi-source monitoring and cross-domain collaboration as claimed in claim 3, characterized in that: The control management module is provided with an industrial computer and a human-computer interaction device; the industrial computer is connected to the electrical components in the electric control cabinet, and the human-computer interaction module communicates with the industrial computer to realize status monitoring, control adjustment, human-computer interaction and operation management.

5. The electric control cabinet operating system for multi-source monitoring and cross-domain collaboration according to claim 1, characterized in that: The server adopts the Hadoop distributed computing framework and utilizes the TensorFlow deep learning framework to build multiple neural network models to achieve fault diagnosis, life prediction and performance evaluation.

6. A multi-source monitoring and cross-domain collaborative electric control cabinet management method, characterized in that: include Real-time detection of various sensor information of the electric control cabinet to monitor the operating status of the electric control cabinet in real time; based on the sensor information, status monitoring, control and adjustment, human-computer interaction and operation management of the electric control cabinet are realized; the sensor information is remotely acquired for big data analysis and fed back to the control management module to realize cross-domain collaboration when the status is abnormal.

7. The electric control cabinet management method of multi-source monitoring and cross-domain collaboration as claimed in claim 6, characterized in that: Remotely obtain the sensor information for big data analysis and feed it back to the control management module when the status is abnormal to achieve cross-domain collaboration, specifically including the following: Acquire real-time infrared image data of the electric control cabinet, and use a pre-trained convolutional neural network to extract key features from the real-time infrared image data to identify the fault type of the electric control cabinet; Acquire real-time sensor information of the electric control cabinet, and input the sensor data into a pre-trained long short-term memory network to predict the remaining service life of the electrical equipment of the electric control cabinet; The performance indicators of the electric control cabinet are obtained, and the performance indicators are input into a pre-trained deep neural network to output a comprehensive performance score of the electric control cabinet.

8. The electric control cabinet management method of multi-source monitoring and cross-domain collaboration as claimed in claim 7, characterized in that: The convolutional neural network includes multiple convolutional layers, pooling layers and fully connected layers; the low-level features in the infrared image data are extracted through the convolutional layers, the feature dimensions are reduced through the pooling layers, and finally the fault type of the electric control cabinet is identified by classification through the fully connected layers; the cross entropy loss function and Adam optimizer are used during the training of the convolutional neural network.

9. The electric control cabinet management method of multi-source monitoring and cross-domain collaboration as claimed in claim 7, characterized in that: The long short-term memory network includes an input layer, multiple LSTM layers and an output layer; the input layer inputs the statistical characteristics of the sensor information and the frequency domain characteristics after Fourier transformation, and the LSTM layer effectively captures the trend of equipment degradation by memorizing long-term dependencies, thereby accurately predicting the remaining service life and outputting it through the output layer; the number of hidden units of the long short-term memory network is set to 128, the mean square error is used as the loss function, and the RMSprop optimizer is used for training.

10. The electric control cabinet management method of multi-source monitoring and cross-domain collaboration as claimed in claim 7, characterized in that: The deep neural network includes an input layer, multiple fully connected layers and an output layer, the input layer inputs a performance indicator; each of the fully connected layers uses a ReLU activation function and unifies the scales of different performance indicators through normalization processing; the output layer uses a Softmax activation function to generate a comprehensive performance score of the electric control cabinet; the cross entropy loss function and Adam optimizer are used during deep neural network training, and the learning rate adopts a learning rate decay strategy.

Images