Low-voltage internet-of-things switch based on cloud platform and edge computing

Abstract

The invention provides a low-voltage internet-of-things switch based on a cloud platform and edge computing, which is connected with a cloud master station, is provided with an edge computing module, and is connected with a plurality of distributed photovoltaic inverters and / or electric vehicle charging piles for centralized management. Parameters needing to be collected are determined by identifying different equipment types, and corresponding control strategies are matched.

Description

Technical Field

[0001] The present invention relates to technical fields such as intelligent power management and control, embedded operating system design, Internet of Things technology, intelligent power equipment, and system safety and stability, and in particular to a low-voltage Internet of Things switch based on a cloud platform and edge computing. Background Art

[0002] With the rapid development of Internet of Things (IoT) technology, the demand for smart home and industrial automation is growing. In low-voltage power systems, the demand for intelligence, remote control and data analysis is increasing. Traditional low-voltage switchgear can no longer meet the requirements of modern power systems for efficient, reliable and intelligent operation. Summary of the invention

[0003] Considering the rapid development of Internet of Things (IoT) technology, traditional low-voltage switchgear can no longer meet the requirements of modern power systems for high efficiency, reliability and intelligence. Therefore, in view of the gaps and deficiencies in the prior art, the present invention proposes a low-voltage IoT switch based on cloud platform and edge computing, which integrates IoT technology, embedded intelligent control module, wireless communication function and cloud master service, can collect and monitor equipment data in real time, and realize accurate optimization management of power equipment through edge computing and cloud master data analysis.

[0004] Its innovation lies mainly in the integrated application of edge computing and cloud master. Through edge computing, IoT switches can monitor device status in real time, process local data, and make intelligent decisions and optimize control without relying on cloud master, which improves the response speed and autonomy of the system. At the same time, cloud master provides remote data management and policy distribution functions, which can accurately control and optimize the equipment, realize cloud-based data analysis and centralized management of large-scale equipment. This combination of edge computing and cloud master significantly improves the intelligence level, operation convenience and overall performance of the system, ensuring efficient and stable operation of equipment in complex power environments.

[0005] The present invention specifically adopts the following technical solutions: A low-voltage IoT switch based on cloud platform and edge computing: connected to a cloud master station and equipped with an edge computing module, connected to a number of distributed photovoltaic inverters and / or electric vehicle charging piles for centralized management; by identifying different device types, the parameters that need to be collected are determined, and the corresponding control strategies are matched.

[0006] Furthermore, the cloud master station sends the control instructions to the IoT switch according to the preset strategy. After receiving the instructions, the IoT switch performs corresponding operations through the edge computing module adaptive edge computing, including adjusting the power output of the photovoltaic inverter or dynamically adjusting the charging rate of the charging pile; the adaptive edge computing automatically adjusts the control strategy based on the real-time data feedback of the equipment operation status.

[0007] Furthermore, the edge computing module applies a dual-path recursive prediction detection algorithm based on the characteristics of the access device to monitor and predict the operating status of the device in real time to determine the parameter change trend including voltage and current, and detect whether there is any abnormal situation.

[0008] Furthermore, the dual-path recursive prediction detection algorithm includes: Route A uses recursive feature matching and neural network model for time series prediction. It uses edge computing to predict the future state of the device by comparing historical records with current data. It monitors whether the current, voltage or power characteristic values ​​of the device are at risk of exceeding the preset threshold. Route B uses a periodic detection method and performs feature analysis after every 2n rounds of data collection. Edge computing analyzes whether the operating status of the equipment is normal through periodic real-time comparison.

[0009] Furthermore, the recursive feature matching mechanism is to implement an iterative process of feature matching by continuously updating and optimizing historical data; the recursive function recurrence_match(current_features, historical_features) is used to compare the feature values ​​of the current input with the historical feature set; each time new data is input, the recursive function is called to determine the similarity between the current data and the historical data, update the historical record, and return the matching result: Maintain a structure to store historical feature data and use indexes to mark important features. When new feature data is input, the recursive function recurrence_match() starts working. The function evaluates the matching degree between the current data and the historical data according to different weights. If the match is successful, it enters the further processing flow. If it is not matched, the recursive process ends. After each recursion, the historical data structure is updated to store the latest feature value; the recursion ends when the current feature successfully matches the historical feature; if all historical features are traversed and no match is found, the recursion ends.

[0010] Furthermore, the calculation process of the recursive feature matching includes: Feature distance calculation: In the recursive process, the difference between the current feature value and the historical feature value is calculated; the similarity between the two feature vectors is calculated and then judged: if the distance value is less than the set threshold, it is determined to be a match, otherwise continue the recursion; Weight allocation: By defining a weight matrix W, it is used to adjust the influence of each feature so that the matching of high-weight features is given priority during recursion.

[0011] Furthermore, in the case of dual-path composite judgment, the judgment results of path A and path B are respectively assigned the same or different weights for comprehensive judgment; During the composite judgment process, the time and number of occurrences of each abnormality are recorded; if the detection results of channel A and / or channel B trigger abnormalities multiple times, the abnormality integral is formed by accumulating the number of abnormalities, and an alarm is triggered when the integral value exceeds the set threshold.

[0012] Furthermore, when an abnormal operation of the device is detected through the dual-path recursive predictive detection algorithm, the IoT switch first attempts to automatically adjust the operating parameters of the device locally to solve the problem based on the adaptive control mechanism; the adaptive control dynamically adjusts parameters including power output or charging rate according to the actual state and characteristics of the device to ensure optimal operation in different environments; if the local adjustment is invalid, the IoT switch triggers an alarm message and uploads the abnormal data to the cloud master station.

[0013] Furthermore, by performing high-frequency sampling on the access devices, characteristic information including current, voltage, power, zero-crossing length and high-order harmonics is extracted. After preprocessing, the recurrent neural network model is trained on the cloud master station, and the trained model is integrated into the TensorFlow Lite Micro model set in the IoT switch MCU; a dual-path recursive prediction detection algorithm is used to perform real-time data monitoring and anomaly prediction.

[0014] Furthermore, the IoT switch communicates with the cloud master station through the Ethernet communication module using a secure communication protocol, establishes a secure communication connection with the charging pile and photovoltaic inverter through the RS485 communication module or the CAN communication module, and performs identity authentication through an encryption protocol to ensure legal access to the device.

[0015] Compared with the prior art, the present invention and its preferred solution can access multiple distributed photovoltaic inverters and electric vehicle charging pile equipment, and realize edge computing and remote control functions through the cloud master station. It can connect to distributed power equipment through standardized interfaces (such as RS485, CAN, Ethernet, etc.) to achieve intelligent centralized management.

[0016] Through edge computing, IoT switches can monitor device status in real time, process local data, and make intelligent decisions and optimize control without relying on the cloud master station, which improves the system's response speed and autonomy. At the same time, the cloud master station provides remote data management and policy distribution functions, which can accurately control and optimize the equipment, realize cloud-based data analysis and centralized management of large-scale equipment. This combination of edge computing and cloud master stations significantly improves the system's intelligence level, ease of operation, and overall performance, ensuring efficient and stable operation of equipment in complex power environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments: Figure 1 This is a scenario architecture diagram of embodiment 1 of the present invention.

[0018] Figure 2 This is a basic module diagram of Example 2 of the present invention.

[0019] Figure 3 This is a schematic diagram of the implementation of the edge computing function of Example 3 of the present invention. DETAILED DESCRIPTION

[0020] Hereinafter, the specific embodiments of the present application will be described in detail with reference to the accompanying drawings. According to these detailed descriptions, those skilled in the art can clearly understand the present application and can implement the present application. Without violating the principles of the present application, the features in the various embodiments can be combined to obtain new implementations, or certain features in certain embodiments can be replaced to obtain other preferred implementations.

[0021] In order to make the features and advantages of this patent more obvious and easy to understand, the following embodiments are specifically described in detail with the accompanying drawings as follows: The embodiment of the present invention proposes a new generation of low-voltage and high-performance IoT switch operating system and cloud master station strategy to realize edge computing design, which can access multiple distributed photovoltaic inverters and electric vehicle charging pile equipment, and realize edge computing and remote control functions through the cloud master station. It is connected to distributed power equipment through standardized interfaces (such as RS485, CAN, Ethernet, etc.) to realize intelligent centralized management.

[0022] Based on the ladder operating system, deep optimization has been carried out to meet the low power consumption requirements of IoT switches, including regional memory allocation, multi-task scheduling and device driver model, to ensure that the system runs efficiently in a multi-tasking environment.

[0023] In terms of data communication, the IoT switch communicates with the cloud master through secure communication protocols (such as TLS-encrypted MQTT, CoAP, etc.), and supports issuing policy instructions from the cloud. Edge computing can dynamically adjust the working status of the device according to the locally processed data by processing and analyzing data in real time on the device side, such as adjusting the power output of the photovoltaic inverter or controlling the charging rate of the charging pile, to achieve optimal utilization of power resources. At the same time, when the cloud master is unavailable, the embedded module can continue to perform basic control functions according to local rules to ensure stable operation of the system.

[0024] 1. System architecture: Hardware layer: It uses an embedded high-performance microcontroller (RISC-V series), supports multi-tasking, has a multi-core main frequency architecture, and integrates NPU (neural processing unit), which can simultaneously control the working status of multiple photovoltaic inverters and charging piles to ensure that the system can still run stably under high load. NPU provides dedicated hardware acceleration for performing machine learning tasks, improving the system's performance in processing complex data and performing intelligent analysis.

[0025] Operating system layer: The operating system layer is a new generation of low-voltage, high-performance operating system, which ensures efficient real-time response and multi-task scheduling of the system, and supports edge computing and remote control functions in complex scenarios. The TensorFlow Lite Micro software package is integrated into the operating system to support the processing and reasoning of machine learning tasks and optimize edge computing capabilities.

[0026] Middleware layer: mainly realizes the unification of communication protocols, data processing and integrated services of equipment management, ensuring efficient interaction between IoT switches and multiple distributed photovoltaic inverters, charging piles and cloud master stations.

[0027] Application layer: provides user interaction and intelligent control. Its main functions include remote control, real-time monitoring, data collection and analysis, and device management and identification.

[0028] It realizes centralized management of multiple distributed photovoltaic inverters and electric vehicle charging piles through standardized interfaces (such as RS485, CAN, Ethernet), supports plug-and-play functions, and ensures the scalability and compatibility of the system by automatically identifying and configuring newly connected devices.

[0029] The IoT switch automatically obtains the device ID and status information by scanning and identifying different photovoltaic inverters and charging piles. Then, it classifies the device according to the type (such as photovoltaic inverter or charging pile) and determines the key parameters that need to be collected. Then, by identifying the device brand, the IoT switch matches the corresponding control algorithm.

[0030] 2. IoT switch operation process: Device startup and initialization: When the IoT switch starts, it self-checks the communication module, power measurement module and storage device to complete hardware initialization. The embedded operating system loads task scheduling, device drivers and communication protocol stacks based on the kernel. The IoT switch establishes a secure communication connection with the charging pile and photovoltaic inverter through Bluetooth or RS485, and authenticates the identity through encryption protocols to ensure legal access to the device.

[0031] Device identification and data collection: The RS485 protocol is used to identify devices and synchronize data, collect parameters such as the power generation and output voltage of the photovoltaic inverter, and the charging status and charging current of the charging pile. These data are stored and cached locally for real-time processing by edge computing and transmission to the cloud master station.

[0032] Edge computing and status monitoring: The IoT switch first identifies the type and brand of the connected power equipment, and applies a dual-path recursive prediction detection algorithm based on the characteristics of the equipment to monitor and predict the operating status of the equipment in real time. Through this algorithm, the system can accurately determine the changing trend of key parameters such as voltage and current, and detect whether there are any abnormal conditions. After that, the IoT switch executes the corresponding optimization algorithm according to the characteristics of the equipment. For example, the system will apply a customized algorithm to adjust the power output or charging rate of the equipment according to different brands of photovoltaic inverters or charging piles to ensure the best operating efficiency. At the same time, the system introduces a dual-path recursive prediction detection algorithm, through which the operating status of the equipment is accurately monitored and predicted. The algorithm can detect the changing trend of parameters such as voltage and current during the operation of the equipment and predict potential abnormal conditions in advance. When abnormalities such as overvoltage or current overload are detected, edge computing combines the prediction results and automatically selects appropriate protection measures according to the characteristics of different equipment brands. This algorithm not only improves the accuracy of status monitoring, but also reduces the dependence on the cloud master station, ensuring that the system can quickly respond to abnormal conditions locally and achieve real-time optimization and protection.

[0033] Abnormal handling and alarm mechanism: When the dual-path recursive predictive detection algorithm detects abnormal operation of the device, the IoT switch will first try to automatically adjust the operating parameters of the device locally to solve the problem according to the adaptive control mechanism. Adaptive control can dynamically adjust key parameters such as power output or charging rate according to the actual status and characteristics of the device to ensure optimal operation in different environments. If the local adjustment is invalid, the IoT switch will trigger an alarm message and upload detailed abnormal data to the cloud master station. The user will receive remote notifications through mobile phone APP and other means, and can intervene remotely. The IoT switch also has automatic recovery and fault handling functions. Through adaptive adjustment and the support of the cloud master station or edge computing, it attempts to restore the normal operation of the device, record the cause of the fault and the handling process, and ensure the stability and reliability of the system.

[0034] Data storage and history: Key equipment data such as power parameters and alarm information are stored locally to facilitate maintenance and fault analysis. The cloud master stores uploaded data for a long time, and users can view historical records and conduct energy efficiency evaluation and optimization through the cloud.

[0035] Equipment maintenance and upgrade: The cloud master supports remote upgrades, and technicians can also perform local debugging and parameter settings via Bluetooth to ensure that the equipment continues to operate efficiently, and update and optimize the configuration based on the analysis results of the cloud master and edge computing.

[0036] 3. Communication modules supported by IoT switches: Ethernet communication module: used to connect the IoT switch to the cloud master station to achieve real-time data transmission and remote control. It updates the device status and power data in real time on the cloud master station and receives remote control instructions to support remote monitoring and management of the device, such as adjusting the inverter working mode or modifying the charging strategy of the charging pile.

[0037] RS485 communication module: used to implement multi-point serial communication in industrial environments and exchange data with photovoltaic inverters and charging pile equipment. It supports long-distance data transmission and multiple device connections, and is suitable for communication scenarios that require high stability and reliability.

[0038] CAN communication module: used to connect and communicate multiple devices, especially in automotive or complex industrial control system environments. Ensure high-speed data transmission between IoT switches and devices such as photovoltaic inverters and charging piles. It is suitable for control systems that require high real-time performance and stability, and ensures the reliability of data transmission.

[0039] 4. Implementation of edge computing: The cloud master sends control instructions to the IoT switch according to the preset strategy. After receiving the instructions, the IoT switch performs corresponding operations through adaptive edge computing, such as adjusting the power output of the photovoltaic inverter or dynamically adjusting the charging rate of the charging pile. The IoT switch adopts an adaptive control algorithm to automatically adjust the control strategy based on the real-time data feedback of the device's operating status. Through the edge computing platform, the system can monitor key parameters such as voltage, current, power, etc. in real time, and dynamically adjust the operating mode of the device according to actual needs. For example, the photovoltaic inverter will automatically optimize the power output under different weather conditions, and the charging pile will adaptively adjust the charging rate according to the battery status of the vehicle to ensure efficient and safe energy management. At the same time, the IoT switch uploads the collected operating data to the cloud master through the communication module for remote monitoring and data analysis, realizing intelligent control and adaptive management of the equipment.

[0040] Preliminary preparation: This includes high-frequency sampling when the electrical device is charging, extracting features such as current, voltage, power, zero-crossing length, and higher harmonics. By processing these features and training the recurrent neural network (RNN) model on the cloud master, the trained model is integrated into the TensorFlow Lite Micro model of the MCU. The ladder operating system supports TensorFlow Lite Micro, which can run the model on the edge device for calculation. Finally, the dual-path recursive prediction detection algorithm is used for real-time data monitoring and anomaly prediction.

[0041] The system periodically scans for connected devices through a standardized communication interface. Once a sub-device is found, the system establishes a connection with it through the device's communication protocol (Modbus, DL / T 645), reads the device's unique ID (MAC address), and identifies the device type and brand.

[0042] The system will adapt the corresponding management strategy and control algorithm according to the type, brand and protocol of the sub-device and list the device in the device management list.

[0043] The overall detection process is monitored by a dual-path recursive prediction detection algorithm and a multi-task balancing mechanism of adaptive control. Each path is responsible for a specific detection task, and finally a composite judgment is made by combining the feature data of the two paths to decide whether to issue a warning or alarm.

[0044] Recursive feature matching mechanism: The core of the recursive feature matching mechanism is to realize the iterative process of feature matching through continuous updating and optimization of historical data. The recursive function recurrence_match(current_features, historical_features) is used to compare the feature value of the current input with the historical feature set. Every time new data is input, the system calls the recursive function to determine the similarity between the current data and the historical data, update the historical record, and return the matching result.

[0045] First, the system maintains a structure to store historical feature data and uses indexes to mark important features. When new feature data such as current, voltage, and power are input, the recursive function recurrence_match() starts working. The function evaluates the degree of match between the current data and the historical data based on different weights (such as the importance of historical features). If the match is successful, further processing is performed; if not, the recursive process ends.

[0046] After each recursion, the system updates the historical data structure and stores the latest feature value in the appropriate location. The recursion ends when the current feature is close enough to the historical feature, and the recursion ends if no match is found after all historical features have been traversed.

[0047] The calculation process of recursive feature matching: Feature distance calculation: During the recursive process, the system calculates the difference between the current feature value and the historical feature value. The similarity between two feature vectors can be calculated using methods such as Euclidean distance and cosine similarity. The calculation formula is as follows: If the distance value is less than the set threshold, it is considered a match, otherwise the recursion continues.

[0048] Weight distribution: Different features may have different importance. Current and voltage have greater weights than power and harmonics. A weight matrix W can be defined to adjust the influence of each feature: In this way, the matching of high-weight features will be given priority during recursion.

[0049] Dual-path recursive prediction detection process: A-Route Process Recursive / memory cursor: Define a data structure to store historical records (current, voltage, power and other data), use timestamps or indexes to mark important feature positions, and the cursor points to these feature values. When new data is input, the cursor is used to determine its match with historical features.

[0050] Current feature analysis: If the currently monitored feature value meets the value condition (>n / m), further processing is performed, otherwise the path is exited.

[0051] RNN (LSTM) model: predicts potential abnormal power usage behavior, collects historical current, voltage and power data, and combines the device-specific power usage pattern as a training set. Use recurrent neural network (RNN) or long short-term memory network (LSTM) to train time series data so that the model can learn and capture the changing patterns of device usage behavior. When new data is input into the model, it can predict and identify abnormal signals such as illegal power usage in real time.

[0052] System feature analysis 1: Continue to monitor feature data, compare historical records, and determine whether feature values ​​fluctuate abnormally.

[0053] Anomaly detection: By comparing historical records, if an abnormality is detected, the situation is recorded and the device status is updated.

[0054] Alarm conditions: If an anomaly is detected multiple times, the system will accumulate the number of anomalies and ultimately decide whether to trigger an alarm based on the alarm conditions.

[0055] Route B process Analyze anomalies every 2n rounds: Road B uses a different strategy and defines a round counter. After every 2n data collections, a large-scale feature analysis is triggered.

[0056] Analysis of real detection value: The system determines whether the current detection value exceeds the safety threshold. If no abnormality is found, the path ends; if the detection value exceeds, the system continues processing.

[0057] System feature analysis 2: Similar to route A, the system enters the feature analysis process and uses the ISTM model to perform complex time series predictions.

[0058] Data cleaning and additional analysis: The system cleans the data of noise and anomalies and performs additional analysis and validation.

[0059] Exception recording and review: Similar to Route A, exceptions on Route B will be recorded and reviewed to ensure accuracy.

[0060] Composite feature screening Dual-path analysis: Path A is high-frequency feature detection, focusing on real-time data monitoring and anomaly prediction; Path B is based on periodic analysis, performing large-scale feature analysis after every 2n rounds of sampling. The combination of the two complements each other through data analysis methods of different frequencies to ensure accuracy.

[0061] Weight mechanism: When making a dual-path composite judgment, different weights can be assigned to the judgment results of path A and path B. At the final decision-making level of the system, the high-frequency data of real-time detection of path A and the results of periodic detection of path B are combined to conduct a comprehensive analysis through the following weighted formula: Comprehensive judgment = WA⋅Path A result + WB⋅Path B result.

[0062] Abnormal accumulation mechanism: During the composite judgment process, the system will record the time and frequency of each abnormal occurrence. If the detection results of channel A or channel B trigger abnormalities multiple times, the system will accumulate these abnormal times to form abnormal integrals, and trigger an alarm when the integral value exceeds the set threshold.

[0063] Time series analysis and multi-layer detection: Through RNN or LSTM models, the system can perform predictive analysis on time series data and capture changes in device usage behavior. The multi-layer screening mechanism in the two paths ensures the accuracy of the detection results, especially for appliance identification and anomaly detection in complex environments.

[0064] When an anomaly is detected and meets the alarm condition, or after several rounds of analysis, no anomaly is found, the detection path is jumped. In the dual-path analysis process, path A mainly provides real-time anomalies through rapid feedback, while path B combines in-depth analysis of historical data. The prediction ability of the LSTM model is integrated into the two paths. By recursively updating historical data, the model continuously optimizes the accuracy of anomaly prediction.

[0065] Adaptive control applications: Application of results from route A: When an abnormality detected by route A requires adjustment, the system automatically optimizes the equipment parameters based on the adaptive control mechanism of route A to ensure the best operating state of the equipment under dynamic load and environmental changes. Adaptive control adjusts equipment parameters based on the real-time detection results of route A to respond to specific abnormal situations.

[0066] Application of B-channel results: The abnormal detection results of B-channel will also trigger adaptive control, and optimize the equipment through periodic large-scale analysis results. The system will adjust the working state of the equipment according to the analysis results of B-channel to avoid failure or overload, thereby improving the overall stability and performance of the equipment.

[0067] System optimization: Adaptive control dynamically adjusts device parameters to ensure that the system maintains optimal performance under different load and environmental conditions. It can respond to changes in the operating status of the device in real time, reduce the probability of failure, and optimize the long-term stability and efficiency of the device. Adaptive control under the multi-task framework enables the optimization schemes and control command tasks of multiple power-consuming devices to be self-balanced, ensuring the optimal allocation and use of computing power and improving the intelligent control capabilities of the system.

[0068] Data upload and monitoring: The adjusted device status and optimization results will be uploaded to the cloud master station through the communication module for remote monitoring and further data analysis. The system will record the adjustment and optimization process to ensure the stability and reliability of the system, and realize intelligent equipment management and efficient operation and maintenance.

[0069] The above solution of the present invention is introduced and demonstrated in more detail through three specific application examples below: Embodiment 1: Figure 1 This is the scenario architecture diagram of the solution. The system includes hardware layer, operating system layer, middleware layer and application layer. The hardware layer includes ARM Cortex-M architecture chip, and the operating system layer is a ladder operating system.

[0070] The IoT switch establishes a connection with the electrical equipment through the CAN bus to achieve one-to-many device scanning. The format is json, and the type of thing is calibrated through the type field.

[0071] A request to read device information is sent to each sub-device through a standard protocol to obtain the device's unique ID, model, brand and other identifiers. PV inverters of different brands may use different data frame structures. The IoT switch will automatically switch to the corresponding protocol parsing module based on the device type.

[0072] Identify the brand and model and match the corresponding control algorithm and management strategy.

[0073] Upload the real-time data of each device to the cloud master in JSON format to achieve real-time monitoring.

[0074] Example 2: IoT switches access multiple distributed photovoltaic inverter devices, and the cloud master station sends policies to realize edge computing functions, such as Figure 2 shown.

[0075] Among them, the device access establishes a communication connection with the photovoltaic inverter using the Modbus protocol via RS485 or CAN bus.

[0076] Read device information: The IoT switch sends an identification request to each connected PV inverter to obtain the device’s unique ID, brand, and device model to ensure successful device identification.

[0077] Device identification: Through RS485 or CAN bus, the IoT switch sends a device identification frame to the connected PV inverter. The frame contains a device query instruction, and the IoT switch requests device ID, brand, model and other information. After receiving the request, the PV inverter returns a response frame containing the unique ID, brand identification and device model of the PV inverter. The response frame also contains the initial status data of the device. The IoT switch parses the returned response frame and determines the specific model and manufacturer of the PV inverter based on the device ID and brand identification.

[0078] Algorithm selection: The IoT switch calls the corresponding scheduling strategy by identifying the brand. For example, for the inverter of brand A, a dynamic maximum power point tracking (MPPT) algorithm may be used to adjust the output power of the inverter in real time. For the inverter of brand B, a voltage limiting algorithm may be required to avoid overload operation.

[0079] Protocol conversion: The core of protocol conversion is to parse the raw data of inverters of different brands into a standard format and convert the control instructions of edge computing back to their respective protocols.

[0080] The IoT switch collects real-time data from connected devices (such as voltage, current, power and other parameters of photovoltaic inverters), which are first processed and stored through local edge computing. The encrypted data is then sent to the cloud master through common IoT communication protocols (such as MQTT, CoAP, HTTP); the cloud master sends the encrypted control instructions to the IoT switch through MQTT, CoAP or HTTP protocols, and the IoT switch converts the brand identified by the device into the corresponding brand protocol through edge computing and sends the instructions to the inverter device.

[0081] Example 3: IoT switches access multiple charging pile devices and cloud master stations to implement edge computing functions, such as Figure 3 As shown: Device access and identification: The IoT switch communicates with multiple charging pile devices through a standardized communication interface. The system regularly scans connected devices, identifies their type and brand by reading the device ID, and adapts the corresponding communication protocol. The IoT switch will assign specific control strategies based on the identified device type, add the device to the management list, and manage it uniformly.

[0082] Cloud master station policy delivery: The IoT switch establishes a connection with the cloud master station through Ethernet, and the cloud master station delivers control policies, such as power allocation, priority setting, or charging policy. After the IoT switch parses the cloud policy, it performs local control, including adjusting the power of the charging pile, distributing the load, and other functions.

[0083] Edge computing: When processing multiple charging pile devices, the IoT switch combines the dual-path (A and B) data analysis framework and uses local edge computing technology to achieve real-time monitoring and exception handling of the device's operating status, reduce dependence on the cloud, and ensure that the device operates efficiently within a safe range.

[0084] A-way process: A-way focuses on the time series prediction system of recursive feature matching and RNN (LSTM) model. By comparing historical records with current data, edge computing is used to predict the future state of the device. Once the current, voltage or power characteristic value of the device is detected to exceed the preset threshold, the IoT switch will immediately start the analysis model at the edge layer to evaluate whether there is abnormal power consumption behavior, and respond quickly to trigger the local alarm mechanism.

[0085] B-way process: B-way uses a periodic detection method and performs an in-depth feature analysis after every 2n rounds of data collection. Edge computing analyzes whether the operating status of the equipment is normal through periodic real-time comparison. At the same time, B-way has powerful data cleaning and review capabilities to ensure more accurate data processing after each detection. If an abnormality is detected, the system will record it locally and continue to track it in the subsequent review process to ensure the stable operation of the equipment.

[0086] The above is a preferred embodiment of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

[0087] This patent is not limited to the above-mentioned optimal implementation mode. Anyone can derive various other forms of a low-voltage IoT switch based on a cloud platform and edge computing under the inspiration of this patent. All equal changes and modifications made according to the scope of the patent application of the present invention should be covered by this patent.

Claims

1. A low-voltage IoT switch based on cloud platform and edge computing, characterized in that: It is connected to the cloud master station and equipped with an edge computing module, and is connected to several distributed photovoltaic inverters and / or electric vehicle charging piles for centralized management. It identifies different device types to determine the parameters that need to be collected and matches the corresponding control strategies.

2. A low-voltage IoT switch based on cloud platform and edge computing according to claim 1, characterized in that: The cloud master sends control instructions to the IoT switch according to the preset strategy. After receiving the instructions, the IoT switch performs corresponding operations through the edge computing module adaptive edge computing, including adjusting the power output of the photovoltaic inverter or dynamically adjusting the charging rate of the charging pile; the adaptive edge computing automatically adjusts the control strategy based on the real-time data feedback of the equipment operation status.

3. A low-voltage IoT switch based on cloud platform and edge computing according to claim 1, characterized in that: The edge computing module applies a dual-path recursive prediction detection algorithm based on the characteristics of the access device to monitor and predict the operating status of the device in real time to determine the parameter change trend including voltage and current, and detect whether there is any abnormal situation.

4. A low-voltage IoT switch based on cloud platform and edge computing according to claim 3, characterized in that: The dual-path recursive prediction detection algorithm includes: Route A uses recursive feature matching and neural network model for time series prediction. It uses edge computing to predict the future state of the device by comparing historical records with current data. It monitors whether the current, voltage or power characteristic values ​​of the device are at risk of exceeding the preset threshold. Route B uses a periodic detection method and performs feature analysis after every 2n rounds of data collection. Edge computing analyzes whether the operating status of the equipment is normal through periodic real-time comparison.

5. A low-voltage IoT switch based on cloud platform and edge computing according to claim 4, characterized in that: The recursive feature matching mechanism is to implement the iterative process of feature matching by continuously updating and optimizing historical data; the recursive function recurrence_match(current_features, historical_features) is used to compare the feature values ​​of the current input with the historical feature set; each time new data is input, the recursive function is called to determine the similarity between the current data and the historical data, update the historical record, and return the matching result: Maintain a structure to store historical feature data and use indexes to mark important features; when new feature data is input, the recursive function recurrence_match() starts working; the function evaluates the matching degree between current data and historical data according to different weights; if the match is successful, enter the further processing flow; If there is no match, the recursive process ends; After each recursion, the historical data structure is updated to store the latest feature values; The recursion ends when the current feature matches the historical feature successfully; If no match is found after all historical features are traversed, the recursion terminates.

6. A low-voltage IoT switch based on cloud platform and edge computing according to claim 5, characterized in that: The calculation process of the recursive feature matching includes: Feature distance calculation: In the recursive process, the difference between the current feature value and the historical feature value is calculated; the similarity between the two feature vectors is calculated and then judged: if the distance value is less than the set threshold, it is determined to be a match, otherwise continue the recursion; Weight allocation: By defining a weight matrix W, it is used to adjust the influence of each feature so that the matching of high-weight features is given priority during recursion.

7. A low-voltage IoT switch based on cloud platform and edge computing according to claim 4, characterized in that: In the case of dual-path composite judgment, the judgment results of path A and path B are assigned the same or different weights for comprehensive judgment; During the composite judgment process, the time and number of occurrences of each abnormality are recorded; if the detection results of channel A and / or channel B trigger abnormalities multiple times, the abnormality integral is formed by accumulating the number of abnormalities, and an alarm is triggered when the integral value exceeds the set threshold.

8. A low-voltage IoT switch based on cloud platform and edge computing according to claim 3, characterized in that: When an abnormal operation of the device is detected through the dual-path recursive predictive detection algorithm, the IoT switch first attempts to automatically adjust the operating parameters of the device locally to solve the problem based on the adaptive control mechanism; the adaptive control dynamically adjusts parameters including power output or charging rate according to the actual status and characteristics of the device to ensure optimal operation in different environments; if the local adjustment is invalid, the IoT switch triggers an alarm message and uploads the abnormal data to the cloud master station.

9. A low-voltage IoT switch based on cloud platform and edge computing according to claim 3, characterized in that: By sampling the access devices at high frequency, we extract characteristic information including current, voltage, power, zero-crossing length and high-order harmonics. After pre-processing, we train the recurrent neural network model on the cloud master station and integrate the trained model into the TensorFlow Lite Micro model set in the MCU of the IoT switch. A dual-path recursive prediction detection algorithm is used for real-time data monitoring and anomaly prediction.

10. A low-voltage IoT switch based on cloud platform and edge computing according to claim 1, characterized in that: The IoT switch communicates with the cloud master station through the Ethernet communication module using a secure communication protocol, establishes a secure communication connection with the charging pile and photovoltaic inverter through the RS485 communication module or the CAN communication module, and performs identity authentication through an encryption protocol to ensure legal access to the device.

Images