Substation power equipment monitoring system and method based on electric Hong Internet-of-Things operating system

Abstract

The invention relates to the technical field of transformer substations, in particular to a transformer substation power equipment monitoring system and method based on an electric Hongyu operating system, and the system comprises a data collection and synchronization module, a frequency domain analysis module, a fault classification optimization module and a real-time early warning monitoring module, and collecting voltage, current and temperature data of the substation equipment. According to the method, the consistency of data is ensured through synchronous processing of real-time data, an accurate basis is provided for subsequent deep analysis, monitoring of abnormal frequency is enhanced by adopting fast Fourier transform, so that the system can quickly identify the power quality problem, and the accuracy of the power quality is improved by combining the support vector machine and the convolutional neural network. According to the method, multiple fault states can be accurately recognized, the fault diagnosis process is optimized, the sensing capacity for tiny performance fluctuation is improved through introduction of the multi-scale entropy, the implementation efficiency of preventive maintenance is improved, and the safety of equipment operation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of substations, and in particular to a substation power equipment monitoring system and method based on a Dianhong IoT operating system. Background Art

[0002] Substations involve the process of receiving, transforming and distributing high-voltage electricity. In substations, high-voltage electricity is reduced or increased in voltage through transformers to meet different transmission or distribution needs. Its core functions include voltage control, power quality management, fault detection and isolation, and system stability assurance. With the advancement of technology, the automation and intelligence level of substations continues to improve, including the use of monitoring systems and automatic control technologies to improve operational efficiency and reliability, reduce human operating errors, and optimize maintenance and operational decisions through real-time data monitoring.

[0003] Among them, the substation power equipment monitoring system is a monitoring and management system designed specifically for power equipment inside the substation. The main purpose of the system is to ensure the efficient and safe operation of power equipment in the substation. By collecting real-time operating data of power equipment, such as voltage, current, and temperature, the monitoring system can promptly detect potential equipment failures or performance degradation, thereby taking preventive measures to avoid large-scale power outages or equipment damage.

[0004] Although existing substation technology includes monitoring and automation functions, it is insufficient in processing and analyzing large amounts of real-time data, resulting in untimely responses to critical faults. The limitations in real-time and data processing capabilities hinder rapid response to complex power events. Fault diagnosis in existing technologies mostly relies on rule-based methods and lacks sufficient flexibility and adaptability to handle unknown or non-standard fault types, resulting in prolonged fault recovery time and increased system unreliability. Existing technologies can only detect equipment performance degradation when it develops to a more serious stage, missing the best time for maintenance and intervention, increasing operating and maintenance costs, and affecting the stability and reliability of power supply. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and to propose a substation power equipment monitoring system and method based on the Dianhong Internet of Things operating system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a substation power equipment monitoring system based on the Dianhong IoT operating system, the system comprising: The data acquisition and synchronization module is based on the power Hongmeng OS, collects the voltage, current and temperature data of substation equipment, performs time synchronization processing on the data, records timestamps and stores them, and organizes them according to the equipment identification and data set time series to generate a power synchronization data set; The frequency domain analysis module uses fast Fourier transform based on the power synchronization data set to analyze the frequency domain characteristics in the power signal, identify abnormal frequency components, and compare the frequency domain characteristics with historical fault data to determine potential abnormal frequencies and generate frequency domain abnormality indicators; The fault classification optimization module classifies and predicts the difference fault states based on the frequency domain abnormality index, uses support vector machines and CNN, optimizes the fault identification process, and generates fault feature diagnosis information; The real-time early warning monitoring module uses multi-scale entropy based on the fault characteristic diagnostic information to evaluate the data change trend, identify early signs of equipment performance degradation, adjust the fault warning threshold according to the entropy value, and trigger an alarm through the power Hongmeng OS when the threshold is exceeded, generating a real-time fault notification signal.

[0007] The present invention is improved in that the steps of acquiring the power synchronization data set are specifically as follows: Based on the power Hongmeng OS, the voltage, current and temperature data of substation equipment are collected and the data time synchronization is processed using the formula:

[0008] Get the adjusted timestamp ,in, is the timestamp of the original data, is the reference timestamp, is the minimum original timestamp, is the adjustment factor; Based on the adjusted timestamp, the data is sorted by device identification and timestamp using the formula:

[0009] Get the sorted data set ,in, represents the original dataset, is the device identifier, is the adjusted timestamp, yes Placeholders in expressions; Based on the sorted data set, data integration is performed using the formula:

[0010] Get the power synchronization data set ,in, is the total number of data points, is the first data points, It is a data point The weight coefficient of Represents the current measurement value of a data point.

[0011] The present invention is improved in that the step of identifying the abnormal frequency component is specifically as follows: Based on the power synchronization data set, the signal data is analyzed in the frequency domain using fast Fourier transform, using the formula:

[0012] Get the converted frequency domain data ,in, is the total number of signal samples, It is The signal value at a time point, is the original time series signal data, is an imaginary unit, is the frequency index; The converted frequency domain data is analyzed to identify abnormal frequency components using the formula:

[0013] Get the abnormal frequency set ,in, is the frequency The amplitude, is the dynamic threshold.

[0014] The present invention is improved in that the step of obtaining the frequency domain abnormality index is specifically as follows: Compare the amplitude of each frequency point in the current frequency domain data with the average amplitude of the corresponding frequency in the historical fault data using the formula:

[0015] Get the abnormal frequency candidate set ,in, and is the frequency in historical data The mean amplitude and standard deviation of is the selectivity factor, is the transformed frequency domain data; Based on the abnormal frequency candidate set, select the frequency that continuously exceeds the historical amplitude threshold, and verify whether the amplitude of the frequency continuously exceeds the target period. If it exceeds the historical amplitude threshold, it is determined to be a potential abnormality, and a potential abnormal frequency set is obtained; Based on the potential abnormal frequency set and the degree of deviation from historical data, the formula is adopted:

[0016] Get frequency domain anomaly indicators ,in, is a verified set of potential anomaly frequencies, is the converted frequency domain data, and is the frequency in historical data The mean amplitude and standard deviation.

[0017] The present invention is improved in that the steps of classification and prediction are specifically as follows: The frequency domain anomaly index is normalized to match the input requirements of the classification model, using the formula:

[0018] Get the standardized index ,in, is the frequency domain anomaly indicator; Based on the standardized indicators, the kernel Extract key features and optimize the accuracy of fault identification using the formula:

[0019] Get the extracted key features ,in, is principal component analysis with applied kernel; Based on the extracted key features, the model is trained for classification using a support vector machine and a convolutional neural network to distinguish different fault states, thereby obtaining a trained classification model for predicting the fault state.

[0020] The present invention is improved in that the step of acquiring the fault characteristic diagnosis information is specifically as follows: Analyze fault data, identify data related to fault characteristics, and use frequency analysis to determine frequency-critical fault types; Based on the frequency-critical fault types, statistical characteristics are extracted, including mean and standard deviation, using the formula: and

[0021] Get the key statistical features, among which, and They are The mean and standard deviation of the dataset, It is a dataset The number of data points in Representation dataset A single data point in .

[0022] The present invention is improved in that the steps of identifying early signs of equipment performance degradation are specifically: Based on the fault characteristic diagnosis information, equipment performance data is collected and sorted into a time series format to obtain equipment performance time series data; Multiscale entropy is applied to evaluate the equipment performance time series data to identify the variability and irregularity of the data, using the formula:

[0023] Obtained at multiple time scales The multi-scale entropy value calculated under ,in, represents the time scale of monitoring, is in scale Next The probability of a state, is the total number of states; A trend analysis is performed on the multi-scale entropy values ​​to determine whether there is a sign of performance degradation.

[0024] The present invention is improved in that the step of obtaining the real-time fault notification signal is specifically as follows: The fault warning threshold is adjusted according to the entropy value, and it is verified whether the current device data exceeds the threshold. If the device data collected in real time exceeds the threshold, an alarm is triggered through the power Hongmeng OS. The judgment formula is:

[0025] Get the alarm trigger status , and generate a real-time fault notification signal, where It is the device data collected in real time. is the adjusted fault warning threshold.

[0026] The substation power equipment monitoring method based on the Dianhong IoT operating system is executed based on the substation power equipment monitoring system based on the Dianhong IoT operating system, and includes the following steps: S1: Based on the power Hongmeng OS, the voltage, current and temperature data of the substation equipment are collected, the data is processed synchronously and the timestamp is recorded, and the data is sorted according to the equipment identification and data time series to generate a power synchronization data set; S2: Based on the power synchronization data set, perform fast Fourier transform, analyze the frequency domain characteristics of the signal, identify abnormal frequency components, and compare historical fault data to determine the abnormal frequency, and generate a frequency domain abnormality index; S3: Based on the frequency domain abnormality index, support vector machine and convolutional neural network are applied to classify and predict the different fault states, optimize the fault identification process, and generate fault feature diagnosis information; S4: Based on the fault feature diagnostic information, multi-scale entropy is used to evaluate the data change trend, identify early signs of equipment performance degradation, adjust the fault warning threshold, and obtain an adjusted threshold; S5: Based on the adjusted threshold, when the threshold is exceeded, an alarm is triggered through the power Hongmeng OS, real-time notification of the fault is implemented, and the device status is continuously monitored, the real-time update of the verification data and the sending of the alarm are verified, and a real-time notification signal of the fault is generated.

[0027] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, the consistency of data is ensured by synchronous processing of real-time data, which provides an accurate basis for subsequent in-depth analysis. The fast Fourier transform is used to enhance the monitoring of abnormal frequency, so that the system can quickly identify power quality problems. The combination of support vector machine and convolutional neural network can accurately identify a variety of fault states and optimize the fault diagnosis process. The introduction of multi-scale entropy improves the perception of small performance fluctuations, enhances the implementation efficiency of preventive maintenance, and improves the safety of equipment operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The present invention proposes a module diagram of a substation power equipment monitoring system based on the Dianhong IoT operating system; Figure 2 A flowchart of obtaining a power synchronization data set in the present invention; Figure 3 is a flow chart for identifying abnormal frequency components in the present invention; Figure 4 This is a flow chart for obtaining frequency domain abnormality indicators in the present invention; Figure 5 A flow chart for classification and prediction in the present invention; Figure 6 This is a flowchart for obtaining fault characteristic diagnosis information in the present invention; Figure 7 A flow chart for identifying early signs of equipment performance degradation in the present invention; Figure 8 This is a flow chart for obtaining the real-time fault notification signal in the present invention. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0030] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined. Example

[0031] See also Figure 1 The present invention provides a technical solution: a substation power equipment monitoring system based on the Dianhong IoT operating system includes: The data acquisition and synchronization module is based on the power Hongmeng OS, collects the voltage, current and temperature data of substation equipment, performs time synchronization processing on the data, records timestamps and stores them, and organizes them according to the equipment identification and data set time series to generate a power synchronization data set; The frequency domain analysis module uses fast Fourier transform based on the power synchronization data set to analyze the frequency domain characteristics in the power signal, identify abnormal frequency components, and compare the frequency domain characteristics with historical fault data to determine potential abnormal frequencies and generate frequency domain abnormality indicators; The fault classification optimization module uses support vector machines and CNN based on frequency domain abnormality indicators to classify and predict differential fault states, optimize the fault identification process, and generate fault feature diagnostic information; The real-time early warning monitoring module uses multi-scale entropy based on fault feature diagnostic information to evaluate data change trends, identify early signs of equipment performance degradation, adjust the fault warning threshold according to the entropy value, and trigger an alarm through the Power Hongmeng OS when the threshold is exceeded, generating a real-time fault notification signal.

[0032] The power synchronization data set includes voltage values, current values, temperature values, and synchronized timestamps. Frequency domain anomaly indicators include peak frequency, amplitude, and signal distortion index. Fault feature diagnostic information includes fault type, fault location, and fault severity. Fault real-time notification signals include notification type, notification time, and notification priority.

[0033] See also Figure 2 ,The specific steps for obtaining the power synchronization data set are: Based on the power Hongmeng OS, the voltage, current and temperature data of substation equipment are collected and the data time synchronization is processed using the formula:

[0034] Get the adjusted timestamp ,in, is the timestamp of the original data, is the reference timestamp, is the minimum original timestamp, It is the adjustment factor, which is used to fine-tune the synchronization accuracy of the timestamp; Based on the adjusted timestamp, the data is sorted by device identification and timestamp using the formula:

[0035] Get the sorted data set ,in, represents the original dataset, is the device identifier, is the adjusted timestamp, yes Placeholders in expressions; Based on the sorted data set, data integration is performed using the formula:

[0036] Get the power synchronization data set ,in, is the total number of data points, is the first data points, It is a data point The weight coefficient of Represents the current measurement value of a data point, such as a voltage, current, or temperature reading.

[0037] Acquire from data acquisition system ; Get from GPS ; Get from the dataset ; set up ; calculate ; "08:00:00" + ("08:00:10 - "07:59:55) * 1.0 = 08:00:00 + 15 seconds * 1.0 = "08:00:15"; use The sorting function, for example, uses Library; The device data includes the timestamps of the three devices. The data before sorting is as follows:

[0038] Applying the sorting formula, the result is:

[0039] Assumptions The data values ​​​​are 100, 200, and 150, respectively , , ; Set to 1.0, 1.2, 0.8 respectively; calculate :

[0040] See also Figure 3 , the specific steps for identifying abnormal frequency components are: Based on the power synchronization data set, the signal data is analyzed in the frequency domain using the fast Fourier transform, using the formula:

[0041] Get the converted frequency domain data ,in, is the total number of signal samples, It is The signal value at a time point, is the original time series signal data, is an imaginary unit, is the frequency index; Analyze the converted frequency domain data and identify abnormal frequency components using the formula:

[0042] Get the abnormal frequency set ,in, is the frequency The amplitude, is the dynamic threshold.

[0043] The collected signal data is: ; ; From 0 to ,consider .

[0044] Calculation process: by For example, calculate :

[0045]

[0046] express The total amplitude of the frequency components is 60, which represents the DC component (zero frequency component) of the time series data.

[0047] The following data is collected:

[0048] , calculated as ; Set to 1.5.

[0049] Calculation process: Evaluate each frequency Whether the abnormal conditions are met: (More than 1.5, so is abnormal) (No more than 1.5, so Not abnormal) (No more than 1.5, so Not abnormal) Abnormal frequency set include ,express The DC component of is abnormally prominent relative to the average amplitude, indicating some abnormal conditions in the power system.

[0050] See also Figure 4 , the specific steps for obtaining frequency domain anomaly indicators are: Compare the amplitude of each frequency point in the current frequency domain data with the average amplitude of the corresponding frequency in the historical fault data using the formula:

[0051] Get the abnormal frequency candidate set ,in, and is the frequency in historical data The mean amplitude and standard deviation of is the selectivity factor used to identify statistically significant deviations, is the transformed frequency domain data; Based on the abnormal frequency candidate set, select the frequency that continuously exceeds the historical amplitude threshold, and verify whether the amplitude of the frequency continuously exceeds the target period. If it exceeds the historical amplitude threshold, it is determined as a potential abnormality, and a potential abnormal frequency set is obtained; Based on the potential abnormal frequency set and the degree of deviation from historical data, the formula is adopted:

[0052] Get frequency domain anomaly indicators ,in, is a verified set of potential anomaly frequencies, is the converted frequency domain data, and is the frequency in historical data The mean amplitude and standard deviation.

[0053] The collected frequencies , then the measured ; set up , ; Calculation process:

[0054] because , so 50Hz is marked as an abnormal frequency, add ; The amplitude of the frequency 50Hz is significantly higher than the historical average, so it is considered a potential abnormal frequency; If a frequency exceeds the threshold in five consecutive measurements, it is considered a potential abnormal frequency. Suppose 50Hz exceeds the threshold four times in the past five measurements, which does not meet the condition.

[0055] Assume that another frequency of 100Hz exceeds the standard for 5 consecutive measurements, and its amplitude is 40dB; set up , ; Amplitude deviation:

[0056] The frequency domain anomaly index calculated for 100 Hz is 14.0625, which means that the degree of anomaly is high, indicating a potential problem with the equipment.

[0057] See also Figure 5 , the steps for classification and prediction are as follows: Normalize the frequency domain anomaly indicators to match the input requirements of the classification model using the formula:

[0058] Get the standardized index ,in, is the frequency domain anomaly indicator; Based on the standardized indicators, kernel PCA is used to extract key features and optimize the accuracy of fault identification. The formula is:

[0059] Get the extracted key features ,in, It is principal component analysis with kernel applied, which is used to capture complex data structures; Based on the extracted key features, the model is classified and trained using support vector machines and convolutional neural networks to distinguish different fault states, and the trained classification model is obtained to predict the fault state.

[0060] The frequency domain abnormality index data obtained from power system monitoring are as follows:

[0061] ; Standardization process:

[0062] It is normalized to the range of 0 to 1, making subsequent feature extraction and machine learning model application easier. More effective.

[0063] The data obtained is [0, 0.25, 0.5, 0.75, 1]; Using the Gaussian kernel to extract the nonlinear relationship in the data, the values ​​of the first two principal components extracted are as follows:

[0064] Through the nuclear Processing standardized data increases the ability to capture complex data patterns and provides input features for the model; The extracted features are input into the model for training to optimize the model’s recognition ability See also Figure 6 , the specific steps for obtaining fault characteristic diagnosis information are as follows: Analyze fault data, identify data related to fault characteristics, and use frequency analysis to determine frequency-critical fault types; Based on the frequency-critical fault types, statistical characteristics are extracted, including the mean and standard deviation, using the formula: and

[0065] Get the key statistical features, among which, and They are The mean and standard deviation of the dataset, It is a dataset The number of data points in Representation dataset A single data point in .

[0066] Capture common failure categories by counting the number of occurrences of each category: Category 1 Occurrences: 4 times Category 2 Occurrences: 5 times Category 3 Occurrences: 3 times If the most common fault category is Category 2, If it is a data set of category 2, the collected data is (measurement of fault severity):

[0067] average value calculate:

[0068] Standard Deviation calculate:

[0069]

[0070] The average value calculated for fault type 2 is 8.6, and the standard deviation calculated for fault type 2 is 1.14, thereby obtaining fault characteristic diagnosis information.

[0071] See also Figure 7 , the steps to identify early signs of equipment degradation are: Based on the fault characteristic diagnosis information, the equipment performance data is collected and sorted into a time series format to obtain the equipment performance time series data; Apply multi-scale entropy to evaluate the equipment performance time series data and identify the variability and irregularity of the data. The formula is:

[0072] Obtained at multiple time scales The multi-scale entropy value calculated under ,in, represents the time scale of monitoring, is in scale Next The probability of a state, is the total number of states; Perform trend analysis on multi-scale entropy values ​​to determine if there are signs of performance degradation.

[0073] At a certain time scale (considering every 2 points as a unit), the data set is as follows:

[0074] Grouping data: In time scale The data is divided into the following groups: [1, 2], [2, 1], [3, 2], [2, 1], [3, 3], [2, 1] Calculating Probability : State [1, 2] appears once, with probability ; State [2, 1] appears 3 times, with probability ; State [3, 2] appears once, with probability ; State [3, 3] appears once, with probability ; Calculate entropy:

[0075] The entropy value obtained It indicates that at a time scale of 2, the complexity and irregularity of the data are high, which is an early sign of device performance degradation.

[0076] See also Figure 8 , the specific steps for obtaining the real-time fault notification signal are as follows: The fault warning threshold is adjusted according to the entropy value, and it is verified whether the current device data exceeds the threshold. If the device data collected in real time exceeds the threshold, an alarm is triggered through the power Hongmeng OS. The judgment formula is:

[0077] Get the alarm trigger status , and generate a real-time fault notification signal, where Real-time collection of equipment data for current status assessment, It is an adjusted fault warning threshold used to detect whether equipment performance falls below safe or expected levels.

[0078] Monitor the temperature of a critical power device where: The latest temperature value directly obtained from the sensor is 102°C; Set as ; Calculation process: The data is collected in real time through sensors on the equipment, and transmitted directly from the sensors to the monitoring system; Real-time comparison of systems : if and , but , triggering an alarm.

[0079] It indicates that the current temperature of the equipment has exceeded the threshold and triggered an alarm, thereby preventing equipment failure or damage and ensuring the safety of equipment and personnel.

[0080] The substation power equipment monitoring method based on the Dianhong IoT operating system includes the following steps: S1: Based on the power Hongmeng OS, the voltage, current and temperature data of the substation equipment are collected, the data is processed synchronously and the timestamp is recorded, and the data is sorted according to the equipment identification and data time series to generate a power synchronization data set; S2: Based on the power synchronization data set, perform fast Fourier transform, analyze the frequency domain characteristics of the signal, identify abnormal frequency components, and compare the historical fault data to determine the abnormal frequency and generate frequency domain abnormality indicators; S3: Based on the frequency domain abnormality indicators, support vector machines and convolutional neural networks are used to classify and predict the different fault states, optimize the fault identification process, and generate fault feature diagnosis information; S4: Based on the fault feature diagnostic information, multi-scale entropy is used to evaluate the data change trend, identify early signs of equipment performance degradation, adjust the fault warning threshold, and obtain the adjusted threshold; S5: Based on the adjusted threshold, when the threshold is exceeded, the Hongmeng OS triggers an alarm, implements real-time notification of faults, and continuously monitors the device status, verifies the real-time update of data and the sending of alarms, and generates a real-time notification signal of faults.

[0081] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. The substation power equipment monitoring system based on Dianhong IoT operating system is characterized by: The system comprises: The data acquisition and synchronization module is based on the power Hongmeng OS, collects the voltage, current and temperature data of substation equipment, performs time synchronization processing on the data, records timestamps and stores them, and organizes them according to the equipment identification and data set time series to generate a power synchronization data set; The frequency domain analysis module uses fast Fourier transform based on the power synchronization data set to analyze the frequency domain characteristics in the power signal, identify abnormal frequency components, and compare the frequency domain characteristics with historical fault data to determine potential abnormal frequencies and generate frequency domain abnormality indicators; The fault classification optimization module classifies and predicts the difference fault states based on the frequency domain abnormality index, uses support vector machines and CNN, optimizes the fault identification process, and generates fault feature diagnosis information; The real-time early warning monitoring module uses multi-scale entropy based on the fault characteristic diagnostic information to evaluate the data change trend, identify early signs of equipment performance degradation, adjust the fault warning threshold according to the entropy value, and trigger an alarm through the power Hongmeng OS when the threshold is exceeded, generating a real-time fault notification signal; The steps for obtaining the power synchronization data set are specifically as follows: Based on the power Hongmeng OS, the voltage, current and temperature data of substation equipment are collected and the data time synchronization is processed using the formula: ; Get the adjusted timestamp ,in, is the timestamp of the original data, is the reference timestamp, is the minimum original timestamp, is the adjustment factor; Based on the adjusted timestamp, the data is sorted by device identification and timestamp using the formula: ； Get the sorted data set ,in, represents the original dataset, is the device identifier, is the adjusted timestamp, yes Placeholders in expressions; Based on the sorted data set, data integration is performed using the formula: ； Get the power synchronization data set ,in, is the total number of data points, is the first data points, It is a data point The weight coefficient of Represents the current measurement value of the data point.

2. According to claim 1, the substation power equipment monitoring system based on the Dianhong IoT operating system is characterized in that: The steps of identifying the abnormal frequency components are specifically as follows: Based on the power synchronization data set, the signal data is analyzed in the frequency domain using fast Fourier transform, using the formula: ； Get the converted frequency domain data ,in, is the total number of signal samples, It is The signal value at a time point, is the original time series signal data, is an imaginary unit, is the frequency index; The converted frequency domain data is analyzed to identify abnormal frequency components using the formula: ； Get the abnormal frequency set ,in, is the frequency The amplitude, is the dynamic threshold.

3. The substation power equipment monitoring system based on the Dianhong IoT operating system according to claim 1 is characterized in that: The steps for obtaining the frequency domain abnormality index are specifically as follows: Compare the amplitude of each frequency point in the current frequency domain data with the average amplitude of the corresponding frequency in the historical fault data using the formula: ； Get the abnormal frequency candidate set ,in, and is the frequency in historical data The mean amplitude and standard deviation of is the selectivity factor, is the transformed frequency domain data; Based on the abnormal frequency candidate set, select the frequency that continuously exceeds the historical amplitude threshold, and verify whether the amplitude of the frequency continuously exceeds the target period. If it exceeds the historical amplitude threshold, it is determined to be a potential abnormality, and a potential abnormal frequency set is obtained; Based on the potential abnormal frequency set and the degree of deviation from historical data, the formula is adopted: ； Get frequency domain anomaly indicators ,in, is a verified set of potential anomaly frequencies, is the converted frequency domain data, and is the frequency in historical data The mean amplitude and standard deviation.

4. The substation power equipment monitoring system based on the Dianhong IoT operating system according to claim 1 is characterized in that: The steps of classification and prediction are specifically as follows: The frequency domain anomaly index is normalized to match the input requirements of the classification model, using the formula: ； Get the standardized index ,in, is the frequency domain anomaly indicator; Based on the standardized indicators, kernel PCA is used to extract key features and optimize the accuracy of fault identification, using the formula: ； Get the extracted key features ,in, is principal component analysis with applied kernel; Based on the extracted key features, the model is trained for classification using a support vector machine and a convolutional neural network to distinguish different fault states, thereby obtaining a trained classification model for predicting the fault state.

5. The substation power equipment monitoring system based on the Dianhong IoT operating system according to claim 1 is characterized in that: The steps for obtaining the fault characteristic diagnosis information are specifically as follows: Analyze fault data, identify data related to fault characteristics, and use frequency analysis to determine frequency-critical fault types; Based on the frequency-critical fault types, statistical characteristics are extracted, including mean and standard deviation, using the formula: and ; Get the key statistical features, among which, and They are The mean and standard deviation of the dataset, It is a dataset The number of data points in Representation dataset A single data point in .

6. The substation power equipment monitoring system based on the Dianhong IoT operating system according to claim 1 is characterized in that: The steps for identifying early signs of equipment performance degradation are as follows: Based on the fault characteristic diagnosis information, equipment performance data is collected and sorted into a time series format to obtain equipment performance time series data; Multiscale entropy is applied to evaluate the equipment performance time series data to identify the variability and irregularity of the data, using the formula: ； Obtained at multiple time scales The multi-scale entropy value calculated under ,in, represents the time scale of monitoring, is in scale Next The probability of a state, is the total number of states; A trend analysis is performed on the multi-scale entropy values ​​to determine whether there is a sign of performance degradation.

7. The substation power equipment monitoring system based on the Dianhong IoT operating system according to claim 1 is characterized in that: The steps for obtaining the real-time fault notification signal are specifically as follows: The fault warning threshold is adjusted according to the entropy value, and it is verified whether the current device data exceeds the threshold. If the device data collected in real time exceeds the threshold, an alarm is triggered through the power Hongmeng OS. The judgment formula is: ； Get the alarm trigger status , and generate a real-time fault notification signal, where It is the device data collected in real time. is the adjusted fault warning threshold.

8. A substation power equipment monitoring method based on the Dianhong IoT operating system is characterized in that: The substation power equipment monitoring system based on the Dianhong IoT operating system according to any one of claims 1 to 8 is implemented, comprising the following steps: Based on the power Hongmeng OS, the voltage, current and temperature data of substation equipment are collected, the data is processed synchronously and the timestamp is recorded, and the data is sorted according to the equipment identification and data time series to generate a power synchronization data set; Based on the power synchronization data set, a fast Fourier transform is performed to analyze the frequency domain characteristics of the signal, identify abnormal frequency components, and compare the historical fault data to determine the abnormal frequency, and generate a frequency domain abnormality index; Based on the frequency domain abnormality index, support vector machine and convolutional neural network are applied to classify and predict the different fault states, optimize the fault identification process, and generate fault feature diagnosis information; Based on the fault characteristic diagnosis information, multi-scale entropy is used to evaluate the data change trend, identify early signs of equipment performance degradation, adjust the fault warning threshold, and obtain an adjusted threshold; Based on the adjusted threshold, when the threshold is exceeded, the Hongmeng OS triggers an alarm, implements real-time notification of faults, continuously monitors the device status, verifies the real-time update of data and the sending of alarms, and generates a real-time notification signal of faults.

Images