Relay protection device state evaluation system and method based on Internet of Things

Abstract

The invention provides a relay protection device state evaluation system and method based on the Internet of Things, and relates to the technical field of the Internet of Things, and the method comprises the steps: obtaining the operation data of a relay protection device in real time through an Internet of Things communication system, the action record and the abnormal alarm information of the relay protection device are acquired; preprocessing the operation data of the relay protection device to obtain processed data; and according to the processed data, establishing an initial relay protection device state evaluation model, and training the relay protection device state evaluation model by using the processed data to obtain a final relay protection device state evaluation model. According to the invention, the data of the relay protection device is acquired and processed in real time through the Internet of Things, the evaluation model is established and trained, accurate evaluation of the state of the device is realized, and the stability and operation and maintenance efficiency of a power system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a relay protection device status evaluation system and method based on the Internet of Things. Background Art

[0002] Traditional condition evaluation methods rely heavily on manual inspections, which are not only time-consuming and laborious, but also limited by the experience and skill level of the inspectors. Human factors may lead to subjectivity and inconsistency in inspection results, affecting the accuracy of condition evaluation. Although regular tests are an important means of evaluating the performance of relay protection devices, these tests are conducted at fixed cycles and may not be able to capture rapid changes in the device status in a timely manner. At the same time, the depth and scope of the test may be limited by resources and time, making it difficult to fully cover all potential problems.

[0003] In the traditional evaluation system, data collection, organization and analysis is a manual and cumbersome process, which may lead to delays in data processing and failure to reflect the actual status of the device in real time. In addition, due to the lack of intelligent data analysis tools and automated monitoring systems, it is difficult to achieve continuous and real-time monitoring of the status of relay protection devices and rapid decision support based on data. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a relay protection device status evaluation system and method based on the Internet of Things, which realizes real-time monitoring and evaluation of the status of the relay protection device.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, a method for evaluating the state of a relay protection device based on the Internet of Things, the method comprising:

[0007] Obtain the operating data of the relay protection device in real time through the IoT communication system, including the operating voltage, current, temperature, switch status parameters, as well as the action records and abnormal alarm information of the relay protection device;

[0008] Preprocessing the operation data of the relay protection device to obtain processed data;

[0009] According to the processed data, an initial relay protection device state evaluation model is established, and the relay protection device state evaluation model is trained using the processed data to obtain a final relay protection device state evaluation model;

[0010] The current operating data of the relay protection device acquired in real time is input into the final relay protection device state evaluation model, the state of the relay protection device is evaluated, and an evaluation result is obtained.

[0011] Furthermore, based on the processed data, an initial relay protection device status evaluation model is established, including:

[0012] Divide the processed data into training set and test set;

[0013] The characteristic data in the processed data is used to construct an initial relay protection device state evaluation model, and the kernel function and initial parameters of the initial relay protection device state evaluation model are determined.

[0014] Furthermore, it also includes: using the training set to train the initial relay protection device state evaluation model to obtain initial training results, and during the training process, real-time monitoring of the performance of the initial relay protection device state evaluation model, the performance including accuracy and loss index;

[0015] According to the initial training results, the initial parameters of the initial relay protection device state evaluation model are adjusted, and the final parameters are determined through repeated experiments and verification;

[0016] The test set is used to evaluate the performance of the initial relay protection device state evaluation model after parameter adjustment, and the performance evaluation results are obtained.

[0017] Furthermore, it also includes: optimizing the initial relay protection device state evaluation model according to the performance evaluation result, and repeatedly training and verifying the optimized initial relay protection device state evaluation model to obtain the final relay protection device state evaluation model.

[0018] Furthermore, the current operating data of the relay protection device obtained in real time is input into the final relay protection device state evaluation model to evaluate the state of the relay protection device and obtain evaluation results, including:

[0019] Processing the current operation data to extract key features for status evaluation, including the voltage fluctuation range, current stability, temperature change trend, and switching action frequency;

[0020] The key features are analyzed, and the current running data acquired each time is assigned to a specific state label to obtain the analysis results.

[0021] Further, the current operating data of the relay protection device acquired in real time is input into the final relay protection device state evaluation model to evaluate the state of the relay protection device and obtain an evaluation result, including: calculating the score of each data point based on the analysis result, and judging the current state of the relay protection device according to the score of each data point to obtain a judgment result;

[0022] According to the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain the evaluation result.

[0023] Furthermore, based on the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain evaluation results, including:

[0024] Summarize the scores of each data point and integrate the judgment results corresponding to each data point;

[0025] Assign corresponding weights to each data point, scoring and judgment result, and calculate the comprehensive evaluation score;

[0026] According to the comprehensive evaluation score, a threshold is set to determine the status of the relay protection device.

[0027] Furthermore, based on the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain evaluation results, including:

[0028] Compare the comprehensive evaluation score with the set threshold to determine the current state of the relay protection device;

[0029] The evaluation results are obtained based on the comprehensive evaluation score and status judgment.

[0030] In the second aspect, a relay protection device status evaluation system based on the Internet of Things includes:

[0031] The data acquisition module is used to collect the operating voltage, current, temperature, switch status operation data of the relay protection device in real time, as well as action records and abnormal alarm information, and transmit the data to the data center;

[0032] The data preprocessing module is used to clean, denoise and normalize the data transmitted by the IoT data acquisition module to obtain processed data;

[0033] The state evaluation model module is used to establish a relay protection device state evaluation model based on the processed data, extract the characteristic values ​​of the relay protection device state, and evaluate the state of the relay protection device based on these characteristic values ​​to obtain an evaluation result;

[0034] The early warning and notification module is used to automatically trigger the early warning mechanism and notify personnel to handle the situation when the relay protection device is in an abnormal state or close to the fault threshold based on the evaluation results of the status evaluation model.

[0035] The above solution of the present invention includes at least the following beneficial effects:

[0036] Through the Internet of Things communication system, this method can obtain the operating data of the relay protection device in real time, including key parameters such as operating voltage, current, temperature, switch status, as well as action records and abnormal alarm information. This realizes comprehensive and real-time monitoring of the relay protection device, which helps to discover and deal with potential problems in a timely manner. The acquired operating data is preprocessed to ensure the accuracy and consistency of the data and provide a high-quality data foundation for model training. Based on the processed data, an initial relay protection device status evaluation model is established, and the final evaluation model is obtained through training. This improves the accuracy and reliability of the status evaluation.

[0037] The current operating data acquired in real time is input into the final status evaluation model to evaluate the status of the relay protection device in real time. This helps to promptly detect abnormal status or potential failures of the device and provide timely early warning information for operation and maintenance personnel. The evaluation results can be used as decision support to help operation and maintenance personnel formulate targeted maintenance strategies and improve maintenance efficiency and effectiveness. Through real-time status evaluation and early warning, this method helps to reduce the occurrence of failures of relay protection devices, thereby reducing the risk of accidents in the power system and improving overall stability. Accurate status evaluation also helps to optimize the configuration and parameter settings of relay protection devices and improve their protection performance and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of a method for evaluating the state of a relay protection device based on the Internet of Things provided by an embodiment of the present invention.

[0039] Figure 2 It is a schematic diagram of a relay protection device status evaluation system based on the Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0041] like Figure 1 As shown, an embodiment of the present invention provides a method for evaluating the state of a relay protection device based on the Internet of Things, and the method comprises the following steps:

[0042] Step 11, obtaining the operating data of the relay protection device in real time through the Internet of Things communication system, including the operating voltage, current, temperature, switch state parameters, and the action record and abnormal alarm information of the relay protection device;

[0043] Step 12, preprocessing the operation data of the relay protection device to obtain processed data;

[0044] Step 13, establishing an initial relay protection device state evaluation model based on the processed data, and training the relay protection device state evaluation model using the processed data to obtain a final relay protection device state evaluation model;

[0045] Step 14: input the current operating data of the relay protection device acquired in real time into the final relay protection device state evaluation model, evaluate the state of the relay protection device, and obtain an evaluation result.

[0046] In an embodiment of the present invention, real-time data monitoring and acquisition can obtain key operating data such as the working voltage, current, temperature, switch status, etc. of the relay protection device in real time and accurately, as well as action records and abnormal alarm information. This helps operation and maintenance personnel to understand the operating status of the device in a timely manner. By preprocessing the acquired operating data, such as data cleaning, format unification, and outlier processing, the quality and availability of the data can be improved. The initial state evaluation model established based on the processed data can obtain a more accurate and reliable final evaluation model after training, thereby improving the accuracy and efficiency of the state evaluation.

[0047] By inputting the current operating data acquired in real time into the final status evaluation model, the status of the relay protection device can be evaluated in real time and accurately. This helps to promptly detect abnormal status or potential faults of the device and trigger the early warning mechanism to provide timely early warning information to operation and maintenance personnel, thereby shortening the fault handling time and reducing the impact of the fault on the power system. The status evaluation results can be used as decision support to help operation and maintenance personnel formulate targeted maintenance strategies, such as regular maintenance and replacement of aging components. This can improve maintenance efficiency and effectiveness, reduce maintenance costs, and extend the service life of the relay protection device. Through real-time status evaluation and early warning, as well as optimized maintenance strategies, this method helps to reduce the occurrence of faults in relay protection devices, thereby reducing the risk of accidents in the power system and improving overall stability and safety.

[0048] In a preferred embodiment of the present invention, the above step 11, obtaining the operating data of the relay protection device in real time through the Internet of Things communication system, including the operating voltage, current, temperature, switch state parameters, and the action record and abnormal alarm information of the relay protection device, may include:

[0049] In an embodiment of the present invention, an Internet of Things communication module (such as a wireless sensor, an RFID tag, etc.) is installed on the relay protection device, and these modules can collect the operating data of the device in real time. An Internet of Things communication gateway is configured to upload the collected data to a cloud server or a local data center. Ensure the network coverage of the Internet of Things communication system, including a wireless network, a wired network, or a hybrid network, to ensure real-time data transmission. The Internet of Things communication module collects data such as the working voltage, current, temperature, and switch state parameters of the relay protection device regularly or according to preset conditions (such as event triggering). The collected data is encapsulated and encrypted by the Internet of Things communication gateway, and then transmitted to the cloud server or the local data center through the network. After receiving the data, the cloud server or the local data center decrypts and parses it to convert the original data into a readable and analyzable format. The processed data is analyzed to identify the features and patterns related to the state of the relay protection device. According to the analysis results, the state of the relay protection device is evaluated. For example, if it is found that the voltage fluctuation is abnormal or the temperature continues to rise, it is judged that the device may be in a warning or fault state.

[0050] Assume that in a smart grid, the IoT communication module collects the working voltage, current, temperature and other data of the relay protection device in real time and uploads the data to the cloud server. The IoT communication module collects the working voltage, current and temperature data of the relay protection device every 5 minutes and uploads it to the cloud server through the wireless network. After receiving the data, the cloud server decrypts and parses it, converts the original data into a readable format, and stores it in the database.

[0051] In a preferred embodiment of the present invention, the above step 12, pre-processing the operation data of the relay protection device to obtain processed data, may include:

[0052] Through data quality checks, missing data caused by sensor failure or communication problems can be identified. For continuously changing data, such as voltage and current, linear interpolation can be used to fill in missing values. Specifically, the value of the missing data point is calculated through a linear equation using the values ​​of two adjacent known data points. The formula for calculating the value of the missing data point is: Where y represents the estimated value of the missing data point; y1 and y2 represent the ordinate values ​​of two adjacent known data points; x1 and x2 represent the abscissa values ​​of two adjacent known data points; and x represents the abscissa of the missing data point. The ordinate value y of the missing data point is estimated by using the values ​​of the two adjacent known data points (x1, y1) and (x2, y2) and the abscissa x of the missing data point.

[0053] Assume that a voltage sensor records voltage values ​​v1 and v3 at time points t1 and t3, respectively, but loses data at time point t2 (t1 < t2 < t3). At this time, the voltage estimate at time t2 is calculated based on the values ​​of v1, v3, t1, t3, and t2 through a linear equation.

[0054] According to the relevant standards and specifications of the power system industry and the actual operation of the relay protection device, set the threshold range. Such as the abnormal value range of voltage and current, the high or low temperature threshold, etc. Traverse the data records, remove the data that exceeds the set threshold range, and record the reasons and time of removal. Identify data from different sources and formats, such as CSV, Excel, etc. Use the minimum-maximum normalization method to convert the numerical range of the data to a specific interval (such as 0-1). Specifically, for each data point, calculate the ratio of the difference between its value and the minimum value to the difference between the maximum value and the minimum value. Traverse the data records to find the minimum and maximum values ​​of each data column. For each data point, calculate its normalized value.

[0055] Set screening criteria related to relay protection device status evaluation. Such as equipment number, time range, data type, etc. Sort the screened data and sort them by time sequence or equipment number and other dimensions. Store the preprocessed data in the server where the evaluation system is located, and set up a regular backup mechanism to prevent data loss. Perform regular maintenance and management of the stored data, including data backup, data recovery, data cleaning and other operations. Establish a data access permission mechanism to ensure that only authorized personnel can access and operate the data.

[0056] In a preferred embodiment of the present invention, the above step 13, establishing an initial relay protection device state evaluation model based on the processed data, may include:

[0057] Step 131, dividing the processed data into a training set and a test set;

[0058] Step 132, using the characteristic data in the processed data, construct an initial relay protection device state evaluation model, and determine the kernel function and initial parameters of the initial relay protection device state evaluation model.

[0059] In an embodiment of the present invention, the preprocessed data set is divided according to a certain ratio (such as 70% as a training set and 30% as a test set). Such a division helps to evaluate the generalization ability of the model, that is, the performance of the model on unseen data. Based on the characteristics of the processed data, a support vector machine (SVM) is selected as a preliminary relay protection device status evaluation model. SVM is a supervised learning model that is good at handling classification and regression problems. Determine the kernel function of the SVM model, such as the radial basis function (RBF). The radial basis function is: K(xc ,x l ) = exp(-γ||x c -x l || 2 )(K(x c ,x l ), where K(x c ,x l ) represents the sample x c and x l γ represents the kernel function value between γ and γ represents the parameter of the radial basis function. It can handle nonlinear relationships in data. Determine the initial parameters of the SVM model, such as learning rate, regularization coefficient, etc. These parameters have an important impact on the performance and training speed of the model and need to be optimized through experiments or cross-validation.

[0060] Assume that a set of operating data of relay protection devices has been acquired and preprocessed, including operating voltage, current, temperature and other characteristics. Here is how to build the initial state evaluation model:

[0061] Assume that the data set contains 1000 records, each with 10 features (including operating voltage, current, temperature, etc.). Divide the data set into 70% training set and 30% test set, i.e., 700 records as training set and 300 records as test set. Use a machine learning library (such as scikit-learn) to build an SVM model. Select RBF as the kernel function and set initial parameters, such as a learning rate of 0.01 and a regularization coefficient of 1.0. Use the training set data to train the SVM model and optimize the performance of the model by adjusting the parameters. By dividing the processed data into training and test sets, it can be ensured that the model will not overfit during the training process, that is, the model will not rely too much on the training data and ignore the overall distribution of the data. In this way, when the model faces new and unseen data, it can still maintain good prediction performance, thereby improving the generalization ability of the model.

[0062] Using the processed feature data to build the initial model and determine the kernel function and initial parameters can make the model more consistent with the distribution and characteristics of the actual data. By continuously adjusting and optimizing these parameters, the model can achieve better convergence during the training process, thereby improving the prediction accuracy and stability of the model. Selecting kernel functions and parameters can make the model more transparent and explainable during the prediction process. For example, some kernel functions can intuitively reflect the similarity and distance between data, thereby helping operation and maintenance personnel to better understand the prediction results and decision-making basis of the model. By building an initial state evaluation model, training and testing it, the state of the relay protection device to be evaluated can be quickly predicted and classified. This helps to shorten the time of state evaluation and improve evaluation efficiency, so as to timely discover and deal with potential problems and ensure the safe and stable operation of the power system.

[0063] In a preferred embodiment of the present invention, the above step 13 further includes:

[0064] Step 134, using the training set to train the initial relay protection device state evaluation model to obtain initial training results, and during the training process, real-time monitoring of the performance of the initial relay protection device state evaluation model, the performance including accuracy and loss index;

[0065] Step 135, adjusting the initial parameters of the initial relay protection device state evaluation model according to the initial training results, and determining the final parameters through repeated tests and verification;

[0066] Step 136, using the test set to perform performance evaluation on the initial relay protection device state evaluation model after parameter adjustment to obtain a performance evaluation result.

[0067] In an embodiment of the present invention, the initial relay protection device state evaluation model is trained using the divided training set data. During the training process, the performance of the initial relay protection device state evaluation model is monitored in real time, including accuracy and loss index. Accuracy refers to the proportion of correct predictions of the model, which reflects the classification ability of the model. The loss index measures the difference between the model prediction results and the actual results, and is used to optimize the training process of the initial relay protection device state evaluation model. According to the initial training results, the initial parameters of the initial relay protection device state evaluation model are adjusted. This includes learning rate, regularization coefficient, kernel function parameters (such as γ value of RBF), etc. Through repeated experiments and verification, the influence of different parameter combinations on model performance is observed, and the parameters are gradually optimized until the final parameter combination is found. The performance of the initial relay protection device state evaluation model after parameter adjustment is evaluated using the test set data. The test set data is data that the model has not seen during the training process, so it can more objectively reflect the generalization ability of the model. By calculating the accuracy and recall rate of the model on the test set, it is evaluated whether the performance of the model meets the actual needs.

[0068] Assume that an initial relay protection device state evaluation model based on SVM has been established, and the training set and test set data have been divided. Use the training set data (70% of the data set) to train the initial relay protection device state evaluation model of SVM. During the training process, the accuracy and loss indicators of the initial relay protection device state evaluation model of SVM are monitored in real time. For example, the accuracy and loss are calculated once for each training epoch, and recorded. The initial parameters are set as: learning rate 0.01, regularization coefficient 1.0, and γ value of RBF kernel function 0.1. Through repeated experiments and verification, it is found that when the learning rate is adjusted to 0.001, the regularization coefficient is adjusted to 0.5, and the γ value is adjusted to 0.5, the performance of the initial relay protection device state evaluation model of SVM reaches the preset effect. Use the test set data (30% of the data set) to evaluate the performance of the initial relay protection device state evaluation model of SVM after parameter adjustment. Calculate the accuracy and recall rate of the initial relay protection device state evaluation model of SVM on the test set. The recall rate calculation formula is: Among them, R represents the recall rate; TP represents the number of correctly classified positive samples; FN represents the number of incorrectly classified negative samples. Assume that the evaluation results are: accuracy 90%, recall 85%. These indicators show that the initial relay protection device status evaluation model of SVM has good performance on the test set and can meet actual needs.

[0069] By using the training set to train the model and monitoring the accuracy and loss indicators in real time, the deficiencies of the model during the training process, such as overfitting and underfitting, can be discovered in time. The model parameters are adjusted according to the initial training results. Through repeated experiments and verification, the parameter combination suitable for the current data can be found, thereby improving the prediction accuracy and generalization ability of the initial relay protection device state evaluation model of SVM. The trained model can quickly and accurately predict and classify the state of the relay protection device to be evaluated, shortening the evaluation time and improving the evaluation efficiency. This helps operation and maintenance personnel to promptly discover and deal with potential problems and ensure the safe and stable operation of the power system. Determining the final parameters through repeated experiments and verification can make the model more stable when facing different data and scenarios, and reduce the degradation of model performance caused by data fluctuations or scenario changes. This helps to improve the reliability and stability of the model in practical applications and provide strong support for the development of smart grids. Accurate performance evaluation results can provide strong decision-making support for operation and maintenance personnel. When the model predicts that a relay protection device may have a risk of failure, the operation and maintenance personnel can take timely measures to overhaul or replace it, thereby avoiding potential safety hazards. Through model training, parameter adjustment and performance evaluation.

[0070] In a preferred embodiment of the present invention, the calculation formula of accuracy is:

[0071]

[0072] Among them, A represents the adjusted accuracy; TP represents the number of relay protection devices correctly predicted by the model to be in abnormal state; TN represents the number of relay protection devices correctly predicted by the model to be in normal state; α and β represent weight adjustment parameters; FP represents the number of normal states incorrectly predicted by the model as abnormal states; FN represents the number of abnormal states incorrectly predicted by the model as normal states; γ represents the adjustment parameter of the total number of predictions.

[0073] In an embodiment of the present invention, (TP+TN)·(1+α) is the weighting of the correct prediction (TP and TN), where α is used to increase the weight of the correct prediction. If α>0, the weight of the correct prediction will increase; if α<0, the weight will decrease. β·(FP+FN) is a penalty term for incorrect predictions. If β>0, it will actually reduce the accuracy (because this is a subtraction operation), but if it is regarded as a negative number (that is, β is actually a penalty coefficient, and it is a negative value or a positive number less than 1 multiplied by a positive number), it increases the penalty for incorrect predictions. Then, consider the adjustment parameter γ of the total number of predictions, 1+γ, which is an adjustment to the total number of predictions (denominator). If γ>0, the total number of predictions will increase; if γ<0, the total number of predictions will decrease (but in actual applications, the total number of predictions will not be reduced, so γ is non-negative). The adjusted accuracy A is obtained by dividing the adjusted numerator (RP+TN)·(1+α)-β·(FP+FN) by the adjusted denominator (TP+TN+FP+FN)·(1+γ).

[0074] Assume that there is a relay protection device prediction model and the following prediction results are obtained:

[0075] TP (number of abnormal states of relay protection devices correctly predicted by the model): 80 units;

[0076] TN (the number of relay protection devices correctly predicted by the model to be in normal state): 60 units;

[0077] FP (number of times the model incorrectly predicts a normal state as an abnormal state): 10 units;

[0078] FN (the number of abnormal states that the model incorrectly predicted as normal): 5 units;

[0079] In addition, assume that the weight adjustment parameters and the total prediction quantity adjustment parameters are as follows:

[0080] α = 0.2 (increase the weight of correct predictions);

[0081] β = -0.3 (penalize wrong predictions, note that this is a negative number);

[0082] γ=0.1 (adjust the total number of predictions). Substitute these values ​​into the formula to calculate the adjusted accuracy Substitute specific values: (Keep four decimal places).

[0083] By introducing the adjustment parameter γ, adaptive adjustments can be made for different data sets or model performance, so that the calculation of accuracy can reflect the actual effect of the model in different scenarios. By analyzing the adjusted accuracy A, it is easier to identify the performance defects of the model in specific categories (abnormal and normal), and then optimize the model in a targeted manner. For example, if the A value is found to be low, the number of FP and FN can be checked to find directions for improvement. The adjustment of accuracy can help decision makers understand the reliability of the model. For example, in power systems, understanding the accuracy of relay protection devices can directly affect the formulation of safety strategies.

[0084] In another preferred embodiment of the present invention, the above step 13 further includes: optimizing the initial relay protection device state evaluation model according to the performance evaluation result, and repeatedly training and verifying the optimized initial relay protection device state evaluation model to obtain the final relay protection device state evaluation model, which may include:

[0085] Carefully review the performance evaluation report, especially key indicators such as accuracy and recall. Develop specific optimization strategies for problems found in the performance evaluation. For example, if the initial relay protection device state evaluation model has a low accuracy in abnormal state prediction, you can consider increasing the proportion of abnormal state samples or adjusting the feature selection of the initial relay protection device state evaluation model. Perform further preprocessing on the training data, such as data cleaning, feature selection, sample balancing, etc. Adjust the parameters of the initial relay protection device state evaluation model (such as learning rate, regularization coefficient, etc.) according to the optimization strategy. Use the adjusted data and parameters to repeatedly train and verify the initial relay protection device state evaluation model. After each training, the model should be evaluated for performance to observe the optimization effect. After multiple iterations and optimizations, when the performance of the initial relay protection device state evaluation model reaches the preset standard, the final relay protection device state evaluation model is determined. Perform a comprehensive performance evaluation of the final model, including its performance under different data sets and scenarios. Deploy the final model to actual applications for relay protection device state evaluation. During use, continuously monitor the performance of the model to promptly identify and address potential problems.

[0086] In a preferred embodiment of the present invention, the above step 14, inputting the current operating data of the relay protection device acquired in real time into the final relay protection device state evaluation model, evaluating the state of the relay protection device, and obtaining the evaluation result, may include:

[0087] Step 114, processing the current operation data to extract key features for status evaluation, wherein the key features include the voltage fluctuation range, current stability, temperature change trend, and switching action frequency;

[0088] Step 115 , analyzing the key features, assigning each acquired current operation data to a specific state tag, and obtaining analysis results.

[0089] In an embodiment of the present invention, sensor types, such as voltage sensors, current sensors, temperature sensors, and switch status sensors, are selected to ensure that they can accurately measure the key parameters of the relay protection device. Configure the data acquisition system and set the data acquisition frequency to ensure that the operating data of the relay protection device can be obtained in real time. Receive the raw data transmitted from the sensor network, perform preliminary data cleaning, and remove obvious outliers and noise, such as sudden interference signals or data transmission errors. Format the data to ensure a unified data format, perform time series analysis on continuous data, and identify trends and periodic changes in the data. Calculate the voltage fluctuation range and fluctuation frequency to evaluate the stability of the voltage. The voltage fluctuation range is obtained by calculating the difference between the maximum and minimum voltage values ​​over a period of time. The formula is: Δv=V max -V min ; Where Δv represents the voltage fluctuation range; V max Indicates the maximum voltage within a time period; V min Indicates the minimum voltage value within a time period. The voltage fluctuation frequency is obtained by analyzing the voltage time series data and counting the number of voltage fluctuations per unit time. This requires filtering and denoising the data to more accurately identify fluctuations. The formula for fluctuation frequency can be expressed as: Where, f represents the voltage fluctuation frequency; N f Indicates the number of voltage fluctuations within a time period T; T represents time. Analyze the smoothness of the current curve, calculate the standard deviation of the current change rate, and identify abnormal current fluctuations, such as overcurrent or undercurrent.

[0090] The smoothness of the current curve is evaluated by calculating the standard deviation of the current change rate. A smooth current curve means that the current change rate is small and the standard deviation is also small. The formula is: Where δ represents the standard deviation of the current change rate; represents the current change rate at the i-th time point; Represents the average value of the current change rate; μ represents the number of data points in a time period. Use time series analysis technology to predict the trend of temperature and identify possible overheating risks, such as continuous temperature rise or exceeding the preset threshold. Count the number of switch actions per unit time, analyze the frequency of switch operations, and evaluate mechanical wear, such as switch contact wear or mechanism jamming. Perform state recognition on the extracted features and establish a mapping relationship between features and states. Set thresholds, for example, the "normal" state score is 0-10 points, the "warning" is 11-50 points, and the "abnormal" is 51-100 points. Assign real-time data to specific state labels such as "normal", "warning", and "abnormal" to facilitate operation and maintenance personnel to quickly understand the status of the device.

[0091] By analyzing key features such as the voltage fluctuation range and current stability in real time, potential fault signs such as abnormal voltage fluctuations and unstable current can be discovered in time, so as to issue early warnings and reduce the occurrence of sudden faults. The status evaluation results can provide strong decision-making support for operation and maintenance personnel. According to the real-time status of the device, operation and maintenance personnel can formulate more reasonable and efficient operation and maintenance plans, such as giving priority to devices with warnings or abnormal conditions, and reducing unnecessary inspections and maintenance work. By continuously monitoring and analyzing the operating status of relay protection devices, problems that may affect the life of the equipment, such as overheating and mechanical wear, can be discovered and handled in time. This helps to extend the service life of the equipment and reduce the cost and frequency of equipment replacement.

[0092] Relay protection devices are an important part of the power system, and the stability and reliability of their status directly affect the operation of the entire power system. Through real-time status evaluation, it can be ensured that the relay protection devices can function normally at critical moments, thereby improving the overall stability of the power system. The real-time status evaluation model analyzes and predicts based on a large amount of historical data and real-time data, providing a more scientific and accurate basis for operation and maintenance decisions. This helps to reduce the subjectivity and uncertainty of human judgment and improve decision-making efficiency and accuracy.

[0093] In a preferred embodiment of the present invention, the above step 14, inputting the current operating state information of the relay protection device obtained in real time into the final relay protection device state evaluation model, evaluating the state of the relay protection device to obtain an evaluation result, also includes:

[0094] Step 144, based on the analysis result, calculating the score of each data point, and judging the current state of the relay protection device according to the score of each data point to obtain a judgment result;

[0095] Step 145, based on the score and judgment result of each data point, comprehensively evaluate the state of the relay protection device to obtain an evaluation result.

[0096] In the embodiment of the present invention, for each data point (such as voltage V, current I, temperature T, switching frequency F, etc.), a score is calculated. This score is determined based on the degree of deviation or abnormality of the data point relative to the normal state. V =f(V); where S V Indicates the voltage rating; V stands for voltage. I =f(I); where S I Indicates the rating of the current; I represents the current. T =f(T); where S T represents the temperature score; T represents the temperature. According to the score of each data point, the current state of the relay protection device can be judged.

[0097] Assume that the real-time operating status information of a relay protection device includes three key characteristics: voltage, current and temperature. Follow the steps below to perform status evaluation:

[0098] Get data on voltage V, current I, and temperature T in real time. Preprocess the data, such as removing noise, normalizing, etc. Set scoring rules for each key feature based on historical data. For example:

[0099] Voltage rating: If V exceeds the normal range (such as ±5%), a corresponding rating will be given based on the degree of excess.

[0100] Current rating: If I exceeds the stable range or reaches the maximum fluctuation range, a rating is given.

[0101] Temperature score: If T is outside the normal range or reaches the overheating threshold, a score is given.

[0102] The scoring formula is used to calculate the score of each data point, and the current status of the relay protection device is determined based on the score of each data point and the preset threshold range. For example:

[0103] If S V ≥Threshold range or S I ≥Threshold range or S T If the score is greater than or equal to the threshold range, the device status is "abnormal". Otherwise, it is judged as "warning" or "normal" according to the proximity of the score.

[0104] In a real-time data acquisition, the voltage V = 225V (exceeding the normal range of 220V), the current I = 10A (within the normal range), and the temperature T = 55°C (close to the overheating threshold of 60°C). According to the scoring rule, S V =2 (because the voltage exceeds the normal range), S I =0 (because the current is within the normal range), S T =1 (because the temperature is close to the overheat threshold).

[0105] By analyzing the score of each data point in real time, it is possible to more accurately predict possible failures of relay protection devices. This precise fault warning helps reduce sudden failures and ensure the stable operation of the power system. Through continuous monitoring and scoring, problems that may affect the life of the device, such as overheating and voltage fluctuations, can be discovered and dealt with in a timely manner. This helps to extend the service life of relay protection devices and reduce the cost of equipment replacement in Shanghai. Relay protection devices are an important part of the power system. By accurately evaluating their status, it can be ensured that the device can function normally at critical moments, thereby improving the reliability of the entire power system. The application of real-time status evaluation models is an important part of the intelligent management of power systems. By evaluating the status of relay protection devices in real time, potential problems can be discovered and dealt with in a timely manner, ensuring the stable operation of the power system, thereby improving user satisfaction.

[0106] In a preferred embodiment of the present invention, the above step 145, based on the score and judgment result of each data point, comprehensively evaluates the state of the relay protection device to obtain an evaluation result, which may include:

[0107] Step 1455, summarizing the scores of each data point and integrating the judgment results corresponding to each data point;

[0108] Step 1456, assign corresponding weights to each data point and scoring and judgment result, and calculate the comprehensive evaluation score;

[0109] Step 1457, according to the comprehensive evaluation score, a threshold is set to determine the state of the relay protection device.

[0110] In the embodiment of the present invention, the score of each data point is summarized: voltage fluctuation score 30; current stability score 50; temperature change score 80. According to the scoring rules, the voltage fluctuation is within the normal range, so the judgment result is "normal". Pay attention to the impact of its score on the comprehensive evaluation, so its status is not directly judged for the time being. Similarly, according to the scoring rules, the current stability is also within the normal range, but its score needs to be considered. The temperature change is close to the overheating threshold, so its score is high and requires special attention.

[0111] Assign corresponding weights to each data point, scoring and judgment result:

[0112] Voltage fluctuation: weight = 0.2 (after normalization);

[0113] Current stability: weight = 0.3 (normalized);

[0114] Temperature change: Weight = 0.5 (after normalization).

[0115] According to the formula S = ∑(R h×W h ) calculates the comprehensive evaluation score; where S represents the comprehensive evaluation score; R h represents the score of the hth data point; W h Represents the weight of the hth data point. Substitute the value S = 30 × 0.2 + 50 × 0.3 + 80 × 0.5 = 6 + 15 + 40 = 61. According to the comprehensive evaluation score, set a threshold to determine the state of the relay protection device: Set the threshold range: normal state (0-40), warning state (41-70), abnormal state (71-100) According to the comprehensive evaluation score of 61, the state of the relay protection device is determined to be "warning state"

[0116] Through real-time monitoring and scoring, potential faults can be discovered in time and the accuracy of early warning can be improved. According to the comprehensive evaluation score, a more reasonable operation and maintenance plan can be formulated, and data points with higher scores can be processed first to improve operation and maintenance efficiency. By timely discovering and dealing with problems that may affect the life of the equipment, such as overheating and voltage fluctuations, the service life of the relay protection device can be extended. The relay protection device is an important part of the power system. Through comprehensive evaluation, it can be ensured that it is in good condition, thereby improving the stability of the entire power system.

[0117] In a preferred embodiment of the present invention, the above step 145, based on the scoring and judgment results of each data point, comprehensively evaluates the state of the relay protection device to obtain an evaluation result, and also includes:

[0118] Step 1458, comparing the comprehensive evaluation score with a set threshold value to determine the current state of the relay protection device;

[0119] Step 1459, determine the evaluation result based on the comprehensive evaluation score and status.

[0120] In an embodiment of the present invention, the threshold range is set: normal state (0-40), warning state (41-70), abnormal state (71-100). The comprehensive evaluation score 61 is compared with the threshold value to determine that the current state of the relay protection device is the "warning state". According to the comprehensive evaluation score and state judgment, the evaluation result is obtained: the relay protection device is currently in the "warning state", and it is necessary to strengthen the monitoring of temperature changes, regularly check the current stability, and pay attention to voltage fluctuations.

[0121] Through real-time monitoring and scoring, as well as the calculation of comprehensive evaluation scores, potential faults can be discovered in a timely manner, and the accuracy and timeliness of early warnings can be improved. This helps operation and maintenance personnel to take quick measures to prevent further expansion of faults and ensure the stable operation of the power system. The comprehensive evaluation score can reflect the overall status of the relay protection device. According to the score, the operation and maintenance personnel can give priority to data points or devices with higher scores and optimize the allocation of operation and maintenance resources. By promptly discovering and dealing with problems that may affect the life of the device, such as overheating and voltage fluctuations, the service life of the relay protection device can be extended.

[0122] like Figure 2 As shown, an embodiment of the present invention further provides a relay protection device status evaluation system 20 based on the Internet of Things, comprising:

[0123] The data acquisition module 21 is used to collect the operating voltage, current, temperature, switch status operation data of the relay protection device, as well as the action record and abnormal alarm information in real time, and transmit the data to the data center;

[0124] The data preprocessing module 22 is used to clean, denoise and normalize the data transmitted by the IoT data acquisition module to obtain processed data;

[0125] The state evaluation model module 23 is used to establish a relay protection device state evaluation model based on the processed data, extract characteristic values ​​of the relay protection device state, and evaluate the state of the relay protection device based on these characteristic values ​​to obtain an evaluation result;

[0126] The early warning and notification module 24 is used to automatically trigger the early warning mechanism and notify personnel to handle the situation when the relay protection device is in an abnormal state or close to the fault threshold according to the evaluation results of the state evaluation model.

[0127] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0128] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0129] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0130] 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 principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for evaluating the state of a relay protection device based on the Internet of Things, characterized in that: The method comprises: Obtain the operating data of the relay protection device in real time through the IoT communication system, including the operating voltage, current, temperature, switch status parameters, as well as the action records and abnormal alarm information of the relay protection device; Preprocessing the operation data of the relay protection device to obtain processed data; According to the processed data, an initial relay protection device state evaluation model is established, and the relay protection device state evaluation model is trained using the processed data to obtain a final relay protection device state evaluation model; The current operating data of the relay protection device acquired in real time is input into the final relay protection device state evaluation model, the state of the relay protection device is evaluated, and an evaluation result is obtained.

2. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 1 is characterized in that: According to the processed data, the initial relay protection device status evaluation model is established, including: Divide the processed data into training set and test set; The characteristic data in the processed data is used to construct an initial relay protection device state evaluation model, and the kernel function and initial parameters of the initial relay protection device state evaluation model are determined.

3. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 2 is characterized in that: Also includes: Using the training set to train the initial relay protection device state evaluation model to obtain initial training results, and during the training process, real-time monitoring of the performance of the initial relay protection device state evaluation model, the performance including accuracy and loss index; According to the initial training results, the initial parameters of the initial relay protection device state evaluation model are adjusted, and the final parameters are determined through repeated experiments and verification; The test set is used to evaluate the performance of the initial relay protection device state evaluation model after parameter adjustment, and the performance evaluation results are obtained.

4. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 3 is characterized in that: Also includes: According to the performance evaluation results, the initial relay protection device state evaluation model is optimized, and the optimized initial relay protection device state evaluation model is repeatedly trained and verified to obtain the final relay protection device state evaluation model.

5. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 4 is characterized in that: The current operating data of the relay protection device obtained in real time is input into the final relay protection device status evaluation model to evaluate the status of the relay protection device and obtain evaluation results, including: Processing the current operation data to extract key features for status evaluation, including the voltage fluctuation range, current stability, temperature change trend, and switching action frequency; The key features are analyzed, and the current running data acquired each time is assigned to a specific state label to obtain the analysis results.

6. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 5 is characterized in that: Inputting the current operating data of the relay protection device obtained in real time into the final relay protection device state evaluation model, evaluating the state of the relay protection device, and obtaining an evaluation result, including: calculating the score of each data point based on the analysis result, and judging the current state of the relay protection device according to the score of each data point to obtain a judgment result; According to the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain the evaluation result.

7. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 6 is characterized in that: According to the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain the evaluation results, including: Summarize the scores of each data point and integrate the judgment results corresponding to each data point; Assign corresponding weights to each data point, scoring and judgment result, and calculate the comprehensive evaluation score; According to the comprehensive evaluation score, a threshold is set to determine the status of the relay protection device.

8. The method for evaluating the state of a relay protection device based on the Internet of Things according to claim 7 is characterized in that: According to the scoring and judgment results of each data point, the status of the relay protection device is comprehensively evaluated to obtain the evaluation results, including: Compare the comprehensive evaluation score with the set threshold to determine the current state of the relay protection device; The evaluation results are obtained based on the comprehensive evaluation score and status judgment.

9. A relay protection device status evaluation system based on the Internet of Things, the system implements the method according to any one of claims 1 to 8, characterized in that: include: The data acquisition module is used to collect the operating voltage, current, temperature, switch status operation data of the relay protection device in real time, as well as action records and abnormal alarm information, and transmit the data to the data center; The data preprocessing module is used to clean, denoise and normalize the data transmitted by the IoT data acquisition module to obtain processed data; The state evaluation model module is used to establish a relay protection device state evaluation model based on the processed data, extract the characteristic values ​​of the relay protection device state, and evaluate the state of the relay protection device based on these characteristic values ​​to obtain an evaluation result; The early warning and notification module is used to automatically trigger the early warning mechanism and notify personnel to handle the situation when the relay protection device is in an abnormal state or close to the fault threshold based on the evaluation results of the status evaluation model.

Images