Intelligent sensor system and method for power grid load monitoring

Abstract

The invention relates to the technical field of power grid monitoring, and discloses an intelligent sensor system and method for power grid load monitoring, and the system comprises a multifunctional sensing module which is used for collecting the electrical parameters and environmental parameters of a power grid in real time, carrying out the preliminary signal processing of the collected data, and transmitting the processed data to a server; the collected data after the preliminary signal processing is output to an edge calculation and anomaly detection module; and the edge calculation and anomaly detection module is used for receiving the collected data of the multifunctional sensing module, performing fusion and compression processing on the collected data, and detecting power grid load anomaly based on rule logic and a machine learning method. Through the innovative design of the multifunctional sensing module, the system can collect electrical parameters and environmental parameters of a power grid in real time, high-precision processing of data is achieved through the signal conditioning and multi-channel fusion technology, the capacity of monitoring the running state of the power grid is remarkably improved, dynamic changes of loads can be comprehensively reflected, and a foundation is laid for precise analysis and regulation.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid monitoring, and particularly to an intelligent sensor system and method for power grid load monitoring. Background Art

[0002] With the development of modern power systems, the operating environment and load characteristics of power grids are becoming increasingly complex. The large-scale access of renewable energy, the rapid popularization of electric vehicles and distributed power sources, and the uncertainty of load at the user side have led to the gradual transformation of traditional power grids towards smart grids. Under this background, the real-time monitoring, accurate analysis, and dynamic optimization of power grid loads have become important links to ensure the safe and stable operation of power grids. However, existing power grid load monitoring technologies face challenges in meeting the complexity and diversity requirements of smart grids, which are specifically manifested as follows: Traditional load monitoring devices mainly focus on single-parameter acquisition, mainly paying attention to basic electrical parameters, and lack the ability to synchronously collect environmental parameters. The influence of environmental factors on the operating state of power grids is becoming increasingly significant, but traditional devices cannot comprehensively consider these factors. In addition, the signal acquisition process is easily affected by electromagnetic interference and noise, resulting in a decline in data quality and the acquisition accuracy not meeting the requirements of modern power grids for high-precision monitoring.

[0003] Existing monitoring systems usually rely on a centralized architecture, uploading the collected data to the cloud for analysis and processing. In scenarios with large amounts of data and long transmission distances, it is easy to cause problems such as high transmission delays and overloading of cloud computing resources. When load anomalies and power grid faults occur, the system cannot quickly identify and respond, missing the best fault handling window and affecting the safety of power grid operation.

[0004] Load monitoring and power grid regulation usually operate separately. The monitoring system mainly provides monitoring data on the load status, while power grid regulation is completed by other systems. Traditional load monitoring systems lack intelligent regulation capabilities based on analysis results and cannot achieve the linkage between load monitoring and regulation, resulting in a lag in the regulation process and poor effects. In addition, the rules of existing regulation systems are single and cannot be dynamically adjusted according to the real-time changes of loads. Especially in the case of a high proportion of distributed energy access, the limitations are more obvious.

[0005] With the continuous expansion of the scale of monitoring data, traditional communication methods face problems such as high bandwidth requirements and low transmission efficiency. Especially in remote monitoring scenarios, the energy consumption of data transmission is too high, restricting the large-scale deployment of the system. At the same time, existing monitoring systems lack effective security encryption mechanisms during data transmission, and the data is vulnerable to tampering and leakage, seriously affecting the security and reliability of the system.

[0006] Therefore, those skilled in the art provide an intelligent sensor system and method for power grid load monitoring to solve the above-mentioned problems. Summary of the Invention

[0007] In view of the deficiencies of the prior art, the present invention provides an intelligent sensor system and method for power grid load monitoring to solve the problems raised in the above background art.

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent sensor system and method for power grid load monitoring, including: A multi-functional sensing module for real-time collection of electrical parameters and environmental parameters of the power grid, and preliminary signal processing of the collected data, and then outputting the collected data after preliminary signal processing to the edge computing and anomaly detection module; An edge computing and anomaly detection module for receiving the collected data of the multi-functional sensing module, performing fusion and compression processing on the collected data, and detecting power grid load anomalies based on rule logic and machine learning methods. The edge computing and anomaly detection module generates a compressed signal after data compression and uploads it to the communication and data transmission module. At the same time, an anomaly event identification signal is generated for rapid response to power grid faults; A communication and data transmission module for receiving the compressed signal and anomaly event identification signal generated by the edge computing and anomaly detection module, and transmitting the data to the cloud analysis and storage module through the local communication network and the remote communication network. The communication and data transmission module supports security encryption during the data transmission process to preferentially transmit key anomaly events to the cloud analysis and storage module; A cloud analysis and storage module for receiving the compressed signal and anomaly event identification signal uploaded by the communication and data transmission module, reconstructing and storing the data, and analyzing the change trend of the power grid load through historical data and real-time data. The cloud analysis and storage module combines anomaly events to trigger a fault diagnosis process, generates a load prediction result and an optimized control instruction, and sends them to the load optimization and feedback control module; A load optimization and feedback control module for receiving the load prediction result and optimized control instruction generated by the cloud analysis and storage module, dynamically adjusting the power grid load distribution. The load optimization and feedback control module executes the optimized control strategy, and feeds back the adjustment result to the multi-functional sensing module and the edge computing and anomaly detection module to real-time adjust the collection range and accuracy, forming a closed loop of monitoring and optimization.

[0009] Preferably, the multi-functional sensing module includes: An electrical quantity monitoring sub-module for collecting voltage , current , and calculating the power factor . The power factor calculation formula is: , where is the voltage value at time t, is the current value at time t, is the instantaneous active power, is the instantaneous apparent power, is the phase angle between the voltage and the current; The signal conditioning sub-module is used to pass the acquired signal through a high-pass filter , a low-pass filter and an anti-aliasing filter to process the acquired signal. The transfer function of the filter is: , where, is the total frequency response function of the signal, indicating the processing effect of each frequency component of the input signal after passing through the system; is the transfer function of the high-pass filter, used to remove low-frequency noise; is the transfer function of the low-pass filter, used to suppress high-frequency interference; is the transfer function of the anti-aliasing filter, used to prevent sampling aliasing.

[0010] Preferably, the edge computing and anomaly detection module includes: The data compression sub-module projects the signal into a low-dimensional space through compressive sensing technology. The mathematical model: , where, represents the fused output data, is the compressed output data, represents the measurement matrix of compressive sensing; The anomaly detection sub-module performs anomaly detection through a rule engine and a support vector machine. Among them, anomalies are detected based on logical rules: , where, represents the load power at time t, with the unit of watt; and represent the minimum and maximum normal values of the load power in sequence.

[0011] Preferably, the edge computing and anomaly detection module fuses the acquired multi-channel signals through a data fusion method. The data fusion result is: , where, represents the data acquired by the i-th sensor, Represents the weight coefficient for data fusion, satisfying ; Represents the output data after fusion.

[0012] Preferably, the communication and data transmission module includes: A local communication sub-module, based on the LoRa protocol, for data transmission between sensor nodes and edge devices. The power consumption calculation formula is: , where Represents the transmission current, with the unit of ampere; Represents the transmission voltage, with the unit of volt; Represents the data transmission time, with the unit of second; A data encryption sub-module, which encrypts the transmitted data using the AES-256 algorithm. The encryption formula: , where Is the encrypted ciphertext, Is the original plaintext, Is the encryption function using the key k.

[0013] Preferably, the cloud analysis and storage module performs load forecasting through a long short-term memory network model. The formula is: , where Is the hidden state at time t; Is the hidden state at time ; And Are the weight matrices between the hidden layer and the input layer; Is the input data; Is the bias term.

[0014] Preferably, the cloud analysis and storage module uses a convolutional neural network to achieve fault classification. The convolution operation formula is: , where Is the eigenvalue of the convolution output; Is the input data; Is the convolution kernel weight; Is the bias term.

[0015] Preferably, the cloud analysis and storage module performs trend forecasting on the power grid load through a time series forecasting method. The time series model for load forecasting includes input historical load data and real-time data.

[0016] Preferably, the load optimization and feedback control module optimizes the power grid load distribution based on the linear programming method, and the optimization objective function is: , wherein, is the load adjustment cost coefficient, is the power of the i-th load.

[0017] An intelligent sensor method for power grid load monitoring includes: Step 1: Collect the electrical parameters and environmental parameters of the power grid in real time. During the collection process, the electrical signals are preliminarily processed through signal conditioning to eliminate noise interference, and the collected data after preliminary signal processing is formed and then used as the output. Step 2: Receive the collected data after preliminary signal processing from Step 1, perform fusion and compression processing on the data to form a fused compressed signal. At the same time, perform anomaly detection on the collected data through rule logic and machine learning methods to generate an anomaly event identification signal, and the compressed signal and the anomaly event identification signal are used as the output. Step 3: Receive the compressed signal and the anomaly event identification signal generated in Step 2, and send the data through the local communication network and the remote communication network. Step 4: Receive the compressed signal and the anomaly event identification signal uploaded in Step 3, reconstruct and store the compressed signal, and analyze the power grid load change trend through historical data and real-time data. According to the anomaly event identification signal, trigger the fault diagnosis process to generate a load prediction result and an optimal control instruction. Step 5: Receive the load prediction result and the optimal control instruction generated in Step 4, dynamically adjust the power grid load distribution based on the optimal control strategy, and feedback the adjusted result to Step 1 and Step 2 to adjust the collection range and accuracy to ensure the continuity of monitoring and optimization, and complete the closed-loop control of the power grid load monitoring.

[0018] The present invention provides an intelligent sensor system and method for power grid load monitoring. It has the following beneficial effects: 1. Through the innovative design of the multi-functional sensing module, the system can collect the electrical parameters and environmental parameters of the power grid in real time, and the signal conditioning and multi-channel fusion technology can achieve high-precision processing of the data, significantly improving the monitoring ability of the power grid operation state, comprehensively reflecting the load dynamic changes, and laying a foundation for accurate analysis and regulation.

[0019] 2. The present invention uses compressed sensing technology and machine learning models to achieve data fusion and compression processing, and quickly detect load anomalies at the edge layer. Compared with the traditional method that relies on cloud analysis, the distributed processing structure reduces the data transmission pressure, improves the response speed of the system to load anomaly events, and provides guarantee for timely early warning and fault location.

[0020] 3. Through the load prediction results and optimized control instructions of cloud analysis, combined with the dynamic regulation algorithm, the present invention adjusts the power grid load distribution in real time, and feeds back the adjustment results to the acquisition module and the edge computing module to achieve the closed-loop linkage of data acquisition, analysis, and regulation, effectively balancing the power grid load and improving the operation stability and energy utilization efficiency.

[0021] 4. The present invention realizes the efficient transmission of data through a multi-protocol communication network, and uses AES encryption technology to ensure data security. The compressed sensing technology further reduces the redundancy of data transmission, reduces communication energy consumption and bandwidth requirements, enables the system to maintain low-power characteristics while achieving efficient transmission, and adapts to complex distributed power grid environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is the system framework diagram of the present invention; Figure 2 is the flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] To enable those skilled in the art to understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0024] The present invention will be described in detail below with reference to the accompanying drawings: Embodiment: Please refer to the attached Figure 1 and the attached Figure 2 , the embodiment of the present invention provides an intelligent sensor system for power grid load monitoring, including: A multi-functional sensing module for real-time collecting the electrical parameters and environmental parameters of the power grid, and performing preliminary signal processing on the collected data, and then outputting the collected data after preliminary signal processing to the edge computing and anomaly detection module; Edge computing and anomaly detection module, which is used to receive the acquisition data from the multi-functional sensing module, fuse and compress the acquisition data, and detect grid load anomalies based on rule logic and machine learning methods. After compressing the data, the edge computing and anomaly detection module generates a compressed signal and uploads it to the communication and data transmission module. Meanwhile, an anomaly event identification signal is generated for a quick response to grid faults; Communication and data transmission module, which is used to receive the compressed signal and the anomaly event identification signal generated by the edge computing and anomaly detection module, and transmit the data to the cloud analysis and storage module through the local communication network and the remote communication network. The communication and data transmission module supports secure encryption during the data transmission process to preferentially transmit critical anomaly events to the cloud analysis and storage module; Cloud analysis and storage module, which is used to receive the compressed signal and the anomaly event identification signal uploaded by the communication and data transmission module, reconstruct and store the data, and analyze the changing trend of the grid load through historical data and real-time data. The cloud analysis and storage module combines anomaly events to trigger a fault diagnosis process, generates a load prediction result and an optimization control instruction, and sends them to the load optimization and feedback control module; Load optimization and feedback control module, which is used to receive the load prediction result and the optimization control instruction generated by the cloud analysis and storage module, dynamically adjust the grid load distribution. The load optimization and feedback control module executes the optimization control strategy, and feeds back the adjustment result to the multi-functional sensing module and the edge computing and anomaly detection module to real-time adjust the acquisition range and accuracy, forming a closed loop of monitoring and optimization.

[0025] Benefits of the multi-functional sensing module: Realize multi-dimensional data acquisition of grid operation, including electrical parameters (voltage, current, power factor) and environmental parameters, comprehensively reflecting the grid operation status; Through the signal conditioning function, effectively reduce the influence of electromagnetic interference and environmental noise during the acquisition process, and improve the accuracy and reliability of the data; The preliminary signal processing function reduces the data processing pressure of subsequent modules and provides high-quality input data for the edge computing and anomaly detection module; Benefits of the edge computing and anomaly detection module: The data fusion function integrates multi-channel acquisition data to form a unified and efficient data structure, reducing data redundancy; The data compression function uses compressive sensing technology to greatly reduce the data transmission volume and relieve the bandwidth pressure of the communication network; The anomaly detection function realizes the rapid identification of load anomalies through rule logic and machine learning methods, helps to detect grid faults in time and issue early warnings, and improves the response speed and operation reliability of the system; Benefits of the communication and data transmission module: Provide multi - protocol support to ensure efficient data transmission under various network conditions; The data encryption function guarantees the data security during transmission, prevents data tampering and leakage, and enhances the overall security of the system; By preferentially transmitting the abnormal event identification signals, it ensures that critical events can be quickly transmitted to the cloud analysis and storage module, shortening the event response time; Benefits of the cloud analysis and storage module: The data reconstruction function restores the compressed signal to its original state, and combined with the storage function, provides a complete data basis for load analysis; The load analysis function combines historical data and real - time data, can accurately predict the changing trend of power grid load, and provides a scientific basis for power grid operation optimization; The fault diagnosis process combines the abnormal event identification signals to quickly identify potential faults and reduce the operation risk of the power grid; The generated load prediction results and optimization control instructions provide reliable decision - making support for subsequent load optimization; Benefits of the load optimization and feedback control module: Based on the load prediction results and optimization control instructions, dynamically adjust the power grid load distribution, improving the efficiency and stability of power grid operation; Through the feedback mechanism, transmit the optimization results to the multi - functional sensing module and the edge computing module, further enhancing the scope and accuracy of data collection, and forming a complete closed - loop control; Realize the linkage of monitoring, analysis and regulation, optimize the power grid resource allocation, reduce energy waste, and at the same time, reduce the operation cost.

[0026] The multi - functional sensing module includes: The electrical quantity monitoring sub - module is used to collect voltage 、current and calculate the power factor The power factor calculation formula is: , where, is the voltage value at time t, is the current value at time t, is the instantaneous active power, is the instantaneous apparent power, is the phase angle between voltage and current; The signal conditioning sub - module is used to process the collected signals through a high - pass filter 、a low - pass filter and an anti - aliasing filter The transfer function of the filter is: , where, is the total frequency response function of the signal, representing the processing effect of each frequency component of the input signal after passing through the system; is the transfer function of the high-pass filter, used to remove low-frequency noise; is the transfer function of the low-pass filter, used to suppress high-frequency interference; is the transfer function of the anti-aliasing filter, used to prevent sampling aliasing.

[0027] Benefits of the electrical quantity monitoring sub-module: It can collect key electrical parameters of the power grid in real time, and comprehensively reflect the operating state of the power grid by calculating the power factor, improving the integrity and applicability of the data; The power factor calculation formula combines voltage, current and phase angle information, can accurately describe the power transmission efficiency of the power grid, helps identify problems of excessive reactive power and phase shift, and provides data support for optimizing the operation of the power grid; Through the dynamic response ability of electrical quantity monitoring, it can adapt to scenarios with drastic load changes, such as new energy access and electric vehicle charging, ensuring the timeliness and accuracy of data collection; Benefits of the signal conditioning sub-module: By removing low-frequency noise with a high-pass filter and suppressing high-frequency interference with a low-pass filter, the signal quality is significantly improved, ensuring the accuracy of the collected data and providing a reliable input for subsequent processing; The anti-aliasing filter effectively prevents information loss and distortion caused by the aliasing effect during signal sampling, ensuring the integrity and reducibility of the signal; The filtering function of the signal conditioning module can adapt to complex electromagnetic environments, avoid performance degradation of the acquisition device due to interference, and provide guarantee for the reliability of power grid operation monitoring in complex environments; Comprehensive benefits: The combination of the electrical quantity monitoring sub-module and the signal conditioning sub-module realizes the full-process improvement from parameter acquisition to signal optimization, the collected data is more accurate, and the anti-interference ability is stronger; The data after signal conditioning is cleaner, providing high-quality data input for the subsequent edge computing and anomaly detection modules, reducing error accumulation, and improving the overall performance of the system; The multi-functional sensing module design fully considers the diversity of the power grid operation environment, and the efficient acquisition and processing capabilities can meet the requirements of modern smart grids for high-precision and multi-dimensional monitoring.

[0028] The edge computing and anomaly detection module includes: The data compression sub-module projects the signal into a low-dimensional space through compressive sensing technology, mathematical model: , Among them, represents the fused output data, is the compressed output data, represents the measurement matrix of compressive sensing; Anomaly detection sub-module, which performs anomaly detection through a rule engine and a support vector machine. Among them, anomalies are detected based on logical rules: , Among them, represents the load power at time t, with the unit of watt; and represent the minimum and maximum normal values of the load power in sequence.

[0029] Benefits of the data compression sub-module: By using compressive sensing technology, high-dimensional data is projected into a low-dimensional space, significantly reducing the data volume. Without significantly reducing the effective information content of the data, it optimizes the data transmission efficiency and alleviates the bandwidth pressure on the communication network; Compressive sensing technology utilizes the sparsity characteristics of signals to maintain the main feature information of the signals while reducing the data dimension, ensuring the accuracy of subsequent analysis and processing; By reducing the transmission overhead through the compressed data, it reduces the requirements for computing resources and storage resources of edge devices and the cloud, and improves the overall operation efficiency of the system; Benefits of the anomaly detection sub-module: The rule engine quickly detects the situation where the load exceeds the normal range based on simple logical rules, and can achieve real-time response, providing guarantee for timely early warning; Adopting a support vector machine model to classify complex features of abnormal events can identify potential abnormal patterns that cannot be covered by the rule engine, further improving the accuracy and adaptability of anomaly detection; By real-time monitoring the power grid load status and quickly locating abnormal points, it can avoid the chain reaction caused by load anomalies and ensure the stability and security of the power grid operation; Comprehensive benefits: The combination of the data compression and anomaly detection modules significantly improves the system's processing ability for large-scale real-time data, reduces the computing burden on the cloud, and improves the overall operation efficiency; Compressive sensing technology optimizes the data stream transmission efficiency, and the combination of the rule engine and machine learning realizes the real-time and accuracy of anomaly detection, providing basic support for power grid fault handling and rapid recovery; In high-load fluctuation scenarios or complex operating conditions, it can meet the high requirements of modern smart grids for data processing and anomaly response through fast compression and accurate detection.

[0030] The edge computing and anomaly detection module fuses the multi-channel signals collected through a data fusion method. The data fusion result is: , where represents the data collected by the i-th sensor, represents the weight coefficient of data fusion, satisfying ; represents the output data after fusion.

[0031] By performing data fusion on multi-channel signals, the collected data from multiple sensors is unified into the output data after fusion, effectively eliminating errors and noises in the single-sensor collection. The fused data can comprehensively reflect the real state of the power grid operation, improving the reliability of data analysis and anomaly detection; The data fusion method uses weight coefficients to perform weighted processing on the importance of data collected by different sensors, making the data contribution match its actual significance and fully utilizing the specialties and advantages of each sensor. Optimize data redundancy and reduce the impact of invalid data on system performance; Multi-channel signal fusion can process multi-dimensional data from different monitoring points in a distributed power grid. By reasonably configuring the weight coefficients, it can adapt to the power grid operation requirements in different scenarios and improve the adaptability of the system under diverse working conditions; The output data after fusion is transmitted as a single data stream to the compression and anomaly detection sub-module, significantly reducing the amount of data that the downstream module needs to process, while ensuring the high quality of the input data, thereby improving the overall processing efficiency of the system and enhancing the effect of compression and anomaly detection.

[0032] The communication and data transmission module includes: The local communication sub-module, based on the LoRa protocol, is used for data transmission between sensor nodes and edge devices. The power consumption calculation formula is: , where represents the transmission current, with the unit of ampere; represents the transmission voltage, with the unit of volt; represents the data transmission time, with the unit of second; The data encryption sub-module uses the AES-256 algorithm to encrypt the transmitted data. The encryption formula: , where is the encrypted ciphertext, is the original plaintext, is an encryption function using the key k.

[0033] Benefits of the local communication sub-module: The local communication sub-module uses the LoRa protocol for data transmission, supports low-power features, accurately quantifies energy consumption with a power consumption calculation formula, and the system optimizes power distribution according to the actual application scenario, effectively extending the operating time of the sensor node; The LoRa protocol has long-distance communication and strong anti-interference capabilities, is suitable for use in the distributed environment of the power grid, can ensure stable and reliable data transmission between the sensor node and the edge device, and can maintain good communication quality in a complex electromagnetic environment; The local communication module does not require complex infrastructure support, is flexible in deployment, and can reduce the overall construction and maintenance costs of the power grid monitoring system; Benefits of the data encryption sub-module: The data encryption sub-module uses the AES-256 algorithm for encryption. This algorithm is an internationally recognized advanced encryption standard, can effectively prevent data from being stolen and tampered with during transmission, and ensure the confidentiality and integrity of the transmitted data; Due to its long key length, the AES-256 encryption algorithm has extremely high anti-attack capabilities, meets the requirements of high security for power grid data transmission, and can guarantee data security in a malicious network environment; The encryption process of the data encryption sub-module is independent of the specific communication protocol, can adapt to different transmission scenarios, including local communication and remote communication, and provides general support for the security of the system; Comprehensive benefits: The combination of local communication and data encryption ensures the transmission efficiency and stability of data, enhances the security of the system, and provides double guarantees for the reliable operation of the power grid monitoring system in a complex environment; The low-power and long-distance characteristics of the local communication sub-module make it suitable for the remote deployment requirements of distributed sensor nodes, while the data encryption function ensures the credibility of data transmission in the distributed system; The communication module achieves a good balance between low-power transmission and high-strength encryption, meets the requirements of the power grid monitoring system for energy efficiency, conforms to the actual needs of high-security transmission, and is suitable for large-scale smart grid application scenarios.

[0034] The cloud analysis and storage module performs load forecasting through a long short-term memory network model, and the formula is: , where is the hidden state at time t; is the hidden state at time ; and is the weight matrix between the hidden layer and the input layer; is the input data; is the bias term; The cloud analysis and storage module uses a convolutional neural network to implement fault classification. The convolution operation formula is: , where, is the eigenvalue of the convolution output; is the input data; is the convolutional kernel weight; is the bias term.

[0035] Benefits of the long short-term memory network model for load forecasting: By memorizing the long-term and short-term load change characteristics, the long short-term memory network model can capture the dynamic fluctuation law of the power grid load, adapt to complex power grid operation scenarios, and improve the accuracy of load forecasting; The special structure of the long short-term memory network model effectively avoids the common gradient vanishing problem in traditional recurrent neural networks, has significant advantages in long time series analysis, and ensures the stability and training effect of the model; The long short-term memory network model can combine real-time input data and historical data for prediction, ensuring that the prediction results can reflect the latest load changes and providing a highly time-sensitive reference for power grid regulation; Benefits of the convolutional neural network for fault classification: The convolutional neural network extracts local features of the input data through convolution operations, and combines multi-layer feature extraction to efficiently identify fault patterns and abnormal features in power grid operation, significantly improving the classification accuracy; The convolution operation can handle the multi-dimensional characteristics of power grid monitoring data. At the same time, it extracts local correlations in time and space, providing a more comprehensive feature identification for complex faults; The convolutional neural network automatically extracts features from the original data, reduces the dependence on manual feature engineering, shortens the development cycle of the fault classification model, and at the same time reduces the dependence on expert experience; Comprehensive benefits: The cloud analysis and storage module combines the long short-term memory network model and the convolutional neural network model, can predict the load change trend, quickly locate the fault cause in case of anomalies, and provide double guarantees for power grid operation; The module uses a large amount of historical data and real-time data, combines intelligent analysis algorithms, realizes seamless connection from data to decision-making, and provides data support for the refined management and optimization of the power grid; Whether it is load forecasting or fault classification, this module can adapt to diverse application requirements in smart grids, including new energy access, electric vehicle load forecasting, and complex fault diagnosis; Accurate load forecasting reduces reactive power waste in power grid operation, and the timeliness and accuracy of fault classification reduce the power outage risk caused by faults, comprehensively improving the operation efficiency and security of the power grid.

[0036] The cloud analysis and storage module predicts the trend of the power grid load through time series prediction methods. The time series model of load forecasting includes input historical load data and real-time data; The load optimization and feedback control module optimizes the power grid load distribution based on the linear programming method. The optimization objective function is: , where is the load adjustment cost coefficient, is the power of the i-th load.

[0037] The cloud analysis and storage module combines historical load data and real-time data through time series prediction methods to achieve accurate prediction of the power grid load change trend, providing a reliable basis for optimized regulation and control; The load optimization and feedback control module is based on the linear programming method, with the load adjustment cost as the objective function, dynamically optimizing the power grid load distribution to ensure the rationality and economy of resource allocation.

[0038] The modules work together to form a closed-loop mechanism from analysis to optimization, improving the operation efficiency and stability of the power grid, enhancing the system's adaptability to complex scenarios, and providing an efficient and scalable solution for the future development of smart grids.

[0039] An intelligent sensor method for power grid load monitoring includes: Step 1: Collect the electrical parameters and environmental parameters of the power grid in real time. During the collection process, the electrical signals are preliminarily processed through signal conditioning to eliminate noise interference, forming the collected data after preliminary signal processing, and then the data is used as the output; Step 2: Receive the collected data after preliminary signal processing from Step 1, perform fusion and compression processing on the data to form a fused compressed signal. At the same time, perform anomaly detection on the collected data through rule logic and machine learning methods to generate an anomaly event identification signal. The compressed signal and the anomaly event identification signal are used as the output; Step 3: Receive the compressed signal and the anomaly event identification signal generated in Step 2, and send the data through the local communication network and the remote communication network; Step 4: Receive the compressed signal and the anomaly event identification signal uploaded in Step 3, reconstruct and store the compressed signal, and analyze the power grid load change trend through historical data and real-time data. According to the anomaly event identification signal, trigger the fault diagnosis process to generate a load forecasting result and an optimized control instruction; Step 5: Receive the load forecasting results and optimized control instructions generated in Step 4, dynamically adjust the power grid load distribution based on the optimized control strategy, and feed back the adjusted results to Steps 1 and 2 for adjusting the acquisition range and accuracy, ensuring the continuity of monitoring and optimization, and completing the closed-loop control of the power grid load monitoring.

[0040] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent sensor system for power grid load monitoring, characterized in that: include: The multifunctional sensing module is used to collect the electrical parameters and environmental parameters of the power grid in real time, perform preliminary signal processing on the collected data, and then output the collected data after preliminary signal processing to the edge computing and anomaly detection module; The edge computing and anomaly detection module is used to receive the collected data from the multifunctional sensing module, fuse and compress the collected data, and detect the abnormality of the power grid load based on rule logic and machine learning methods. The edge computing and anomaly detection module compresses the data to generate a compressed signal and uploads it to the communication and data transmission module. At the same time, it generates an abnormal event identification signal for rapid response to power grid failures; The communication and data transmission module is used to receive the compressed signal and abnormal event identification signal generated by the edge computing and abnormality detection module, and transmit the data to the cloud analysis and storage module through the local communication network and the remote communication network. The communication and data transmission module supports secure encryption during the data transmission process to give priority to transmitting key abnormal events to the cloud analysis and storage module; The cloud analysis and storage module is used to receive the compressed signal and abnormal event identification signal uploaded by the communication and data transmission module, reconstruct and store the data, and analyze the change trend of the power grid load through historical data and real-time data. The cloud analysis and storage module triggers the fault diagnosis process in combination with the abnormal event, generates the load prediction results and optimization control instructions, and sends them to the load optimization and feedback control module; The load optimization and feedback control module is used to receive the load forecast results and optimization control instructions generated by the cloud analysis and storage module, and dynamically adjust the load distribution of the power grid. The load optimization and feedback control module executes the optimization control strategy and feeds back the adjustment results to the multi-functional perception module and the edge computing and anomaly detection module to adjust the collection range and accuracy in real time, forming a closed loop of monitoring and optimization.

2. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The multifunctional sensing module comprises: Electrical quantity monitoring submodule, used to collect voltage , Current , and calculate the power factor , the power factor calculation formula is: ， in, is the voltage value at time t, is the current value at time t, is the instantaneous active power, is the instantaneous apparent power, is the phase angle between voltage and current; Signal conditioning submodule for high-pass filtering , low pass filter and anti-aliasing filters The collected signal is processed and the filter transfer function is: ， in, It is the total frequency response function of the signal, which indicates the processing effect of the input signal on each frequency component after passing through the system; is the transfer function of the high-pass filter, which is used to remove low-frequency noise; is the transfer function of the low-pass filter, which is used to suppress high-frequency interference; is the transfer function of the anti-aliasing filter, which is used to prevent sampling aliasing.

3. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The edge computing and anomaly detection module includes: The data compression submodule projects the signal into a low-dimensional space through compressed sensing technology. The mathematical model is: ， in, represents the fused output data, is the compressed output data, Represents the measurement matrix of compressed sensing; The anomaly detection submodule performs anomaly detection through the rule engine and support vector machine, where anomalies are detected based on logical rules: ， in, represents the load power at time t, in watts; and Indicates the minimum and maximum normal values ​​of load power respectively.

4. The intelligent sensor system for power grid load monitoring according to claim 3, characterized in that: The edge computing and anomaly detection module fuses the collected multi-channel signals through a data fusion method, and the data fusion result is: ， in, represents the data collected by the i-th sensor, Represents the weight coefficient of data fusion, satisfying ; Represents the fused output data.

5. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The communication and data transmission module comprises: The local communication submodule is based on the LoRa protocol and is used for data transmission between sensor nodes and edge devices. The power consumption calculation formula is: ， in, Indicates the transmission current in amperes; Indicates the transmission voltage in volts; Indicates the data transmission time in seconds; The data encryption submodule uses the AES-256 algorithm to encrypt the transmitted data. The encryption formula is: ， in, is the encrypted ciphertext, is the original plaintext, is an encryption function using key k.

6. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The cloud analysis and storage module performs load forecasting through a long short-term memory network model, and the formula is: ， in, is the hidden state at time t; It's time The hidden state of and is the weight matrix of the hidden layer and the input layer; is the input data; is the bias term.

7. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The cloud analysis and storage module uses a convolutional neural network to implement fault classification. The convolution operation formula is: ， in, is the eigenvalue of the convolution output; is the input data; is the convolution kernel weight; is the bias term.

8. The intelligent sensor system for power grid load monitoring according to claim 1, characterized in that: The cloud analysis and storage module predicts the trend of power grid load through a time series prediction method, and the time series model of load prediction includes inputting historical load data and real-time data.

9. The intelligent sensor system for power grid load monitoring according to claim 8, characterized in that: The load optimization and feedback control module optimizes the grid load distribution based on the linear programming method, and the optimization objective function is: ， in, is the load adjustment cost coefficient, is the power of the ith load.

10. An intelligent sensor method for power grid load monitoring, according to an intelligent sensor system for power grid load monitoring according to any one of claims 1 to 9, characterized in that: include: Step 1: Real-time collection of electrical parameters and environmental parameters of the power grid. During the collection process, the electrical signal is preliminarily processed by signal conditioning to eliminate noise interference, forming the collected data after preliminary signal processing, and then the data is used as output; Step 2: receiving the collected data after the preliminary signal processing from step 1, fusing and compressing the data to form a fused compressed signal; at the same time, performing anomaly detection on the collected data through rule logic and machine learning methods to generate an abnormal event identification signal, and the compressed signal and the abnormal event identification signal are used as output; Step 3, receiving the compressed signal and abnormal event identification signal generated in step 2, and sending the data through the local communication network and the remote communication network; Step 4: Receive the compressed signal and abnormal event identification signal uploaded in step 3, reconstruct and process the compressed signal and store it, analyze the load change trend of the power grid through historical data and real-time data, trigger the fault diagnosis process according to the abnormal event identification signal, and generate load prediction results and optimization control instructions; Step 5: Receive the load forecast results and optimization control instructions generated in step 4, dynamically adjust the grid load distribution based on the optimization control strategy, and feed back the adjusted results to steps 1 and 2 to adjust the acquisition range and accuracy, ensure the continuity of monitoring and optimization, and complete the closed-loop control of grid load monitoring.

Images