Electrical power equipment fault monitoring method and system based on industrial internet

Abstract

The invention discloses an electrical power equipment fault monitoring method and system based on the industrial internet, and relates to the technical field of electrical equipment fault monitoring, and the method comprises the steps: collecting the power related data of electrical equipment, constructing a power related data set H, calculating the discharge intensity change rate [delta] p, rapidly recognizing abnormal electrical equipment, constructing an abnormal equipment group, and then, carrying out the fault monitoring of the abnormal electrical equipment through the collection of the power related data, construction of a power related data set H and calculation of the discharge intensity change rate [delta] p. The method comprises the following steps: extracting characteristic indexes of each abnormal electrical device in an abnormal device group, predicting the device aging rate vaming of each electrical device in combination with a linear regression algorithm, calculating to obtain the device remaining service life RUL of the electrical device, and finally constructing a device comprehensive fault score Gz, performing risk grade division, and triggering an intelligent early warning mechanism. The maintenance decision of the electrical equipment is optimized, the risk of fault occurrence is reduced, and the stability and safety of a power system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment fault monitoring, and specifically to an electrical power equipment fault monitoring method and system based on the Industrial Internet. Background Art

[0002] With the development of the Industrial Internet, many industries, especially the power and electrical fields, are gradually moving towards intelligent and automated transformation. The Industrial Internet has promoted the establishment of systems such as equipment health monitoring, remote control and fault warning by combining various devices and sensors with cloud computing, big data, artificial intelligence and other technologies. In the power system, the health status of the equipment is directly related to the stable operation and safety of the power system, especially electrical equipment such as transformers, cables and power switches. The operation of electrical equipment determines the reliability and safety of the power grid. For electrical equipment, the fault monitoring system has become an important tool to ensure the stable operation of the power system. Through intelligent sensor groups and data analysis methods, the status of electrical equipment can be monitored in real time, and potential failure risks of equipment can be quickly discovered, so as to make timely maintenance or replacement decisions.

[0003] Currently, fault monitoring of electrical equipment mainly relies on traditional physical inspections and regular maintenance methods, such as regular checks of insulation resistance, partial discharge, temperature and other parameters. Although these methods can provide a preliminary assessment of the equipment's operating status, they have some shortcomings, such as reliance on manual inspections and delayed data collection. Furthermore, when dealing with complex and aging equipment after long-term use, it is difficult to make effective maintenance decisions in advance, which increases the risk of equipment failure and downtime.

[0004] Traditional monitoring systems lack intelligent analysis and prediction capabilities, making it difficult to comprehensively assess the health status and predict aging of equipment based on historical operating data. This lack of intelligent prediction leads to operations and maintenance personnel typically handling equipment failures after they occur, making it difficult to prevent equipment degradation in advance based on failure modes, resulting in missed maintenance opportunities. These problems ultimately increase the maintenance costs of power companies in equipment management and even affect the normal power supply of the grid, resulting in safety hazards and economic losses. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides an electrical power equipment fault monitoring method and system based on the industrial Internet, which solves the problems in the above-mentioned background technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an electrical power equipment fault monitoring system based on the Industrial Internet, including a data acquisition and preprocessing module, a data analysis module, a feature extraction module, an aging model prediction module, and an intelligent early warning feedback module;

[0007] The data acquisition and preprocessing module is used to collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center;

[0008] The data analysis module is used to calculate the discharge intensity change rate Δp based on the power-related data set H, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group;

[0009] The feature extraction module is used to extract the characteristic indicators of each electrical device in the abnormal electrical equipment group and construct a feature vector G;

[0010] The aging model prediction module is used to build a device aging rate prediction model and obtain the device aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment;

[0011] The intelligent early warning feedback module is used to conduct risk assessment on electrical equipment based on the equipment comprehensive fault score Gz, classify the fault risk level of electrical equipment, and trigger the intelligent early warning mechanism.

[0012] Preferably, the data acquisition and preprocessing module includes a data acquisition unit and a data processing unit;

[0013] The data acquisition unit is used to deploy smart sensor groups on electrical equipment and use the smart sensor groups to collect power-related data generated during the operation of the power equipment in real time, where the power-related data includes the cumulative usage time t used , electrical equipment temperature T, electrical equipment total current I total , voltage V, insulation resistance r ins and partial discharge intensity p discharge ;

[0014] The intelligent sensor group includes temperature sensor, current monitoring sensor, voltage sensor, insulation resistance sensor and partial discharge sensor;

[0015] The real-time collected power-related data is transmitted to the local server via wireless Wi-Fi technology.

[0016] Preferably, the data processing unit is used to process the power-related data in the local server, including anomaly detection and correction, denoising and missing value filling;

[0017] Anomaly detection and correction refers to using the isolation forest algorithm to detect abnormal data points in the data and correct the outliers based on the neighboring data;

[0018] Denoising refers to using wavelet transform denoising method to decompose the noise in the data into different frequency components and remove the high-frequency noise components;

[0019] Missing value filling refers to filling missing values in the time series data of equipment operation status using spline interpolation method;

[0020] Based on the processed power-related data, a power-related data set H is constructed and stored in a power data center, wherein the power-related data set H includes power-related data generated by several electrical devices during operation.

[0021] Preferably, the data analysis module is used to analyze the operating conditions of the electrical equipment over a period of time based on the power-related data set H, and calculate the discharge intensity change rate Δp. The specific calculation process of the discharge intensity change rate Δp is:

[0022]

[0023] Where p discharge (t) represents the partial discharge intensity of the electrical equipment at the current time point t, p discharge (t-1) represents the partial discharge intensity of the electrical equipment at the previous time point t-1;

[0024] A partial discharge change threshold Fd is preset, and the discharge intensity change rate Δp is compared with the partial discharge change threshold Fd. When the discharge intensity change rate Δp is greater than or equal to the partial discharge change threshold Fd, that is, Δp ≥ Fd, the early warning mechanism is automatically triggered. At this time, the electrical equipment is judged to be in an abnormal state, the electrical equipment in the abnormal state is marked, and an abnormal electrical equipment group is constructed.

[0025] Preferably, the feature extraction module is used to obtain the power-related data of each abnormal electrical device in the abnormal electrical device group according to the power-related data set H, and extract the characteristic indicators of each abnormal electrical device, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ);

[0026] Analyze the relationship between temperature and discharge intensity to obtain the temperature-discharge intensity correlation factor Xg p，T , where the temperature-discharge intensity correlation factor Xg p，T The specific way to obtain it is:

[0027]

[0028] In the formula, n represents the total number of data sampling times, p discharge (t i ) represents the time point ti The partial discharge intensity at represents the mean value of the local discharge intensity, T(t i ) represents the time point t i The temperature of the electrical equipment at represents the average temperature of the electrical equipment, i represents the data sampling index, i = {1, 2, 3, ..., n}.

[0029] Preferably, the insulation resistance fluctuation amplitude between consecutive time points is analyzed to obtain the insulation resistance fluctuation coefficient JY, wherein the insulation resistance fluctuation coefficient JY is specifically obtained as follows:

[0030]

[0031] Where r ins (t i ) represents the time point t i The insulation resistance of electrical equipment, r ins (t i-1 ) Time point t i-1 Insulation resistance of electrical equipment;

[0032] Analyze the power loss of the insulation material of the electrical equipment to calculate the dielectric loss factor tan(δ). The dielectric loss factor tan(δ) is obtained as follows:

[0033]

[0034] Where V represents voltage, I total It represents the total current of the electrical equipment, cos(φ) represents the phase difference between the voltage and current waveforms, R reactive represents the reaction resistance;

[0035] Based on the extracted characteristic indicators, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ), construct a characteristic vector G, where the characteristic vector G includes characteristic indicators of each electrical device.

[0036] Preferably, the aging model prediction module includes a model building unit and a prediction unit;

[0037] The model building unit is used to build an equipment aging rate prediction model using a linear regression algorithm and obtain historical power-related data of each abnormal electrical equipment in the power data center. The specific form of the equipment aging rate prediction model is:

[0038] v aming =ω1·Xg p，T +ω2·JY+ω3·tan(δ)+b;

[0039] Where ω1, ω2 and ω3 represent the temperature-discharge intensity correlation factor Xg p，T , the regression coefficients of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), b represents the bias term;

[0040] Feature extraction is performed on the historical power-related data of each abnormal electrical equipment, and the extracted feature indicators are input into the equipment aging rate prediction model for model training. The mean square error (MSE) is used as the loss function to minimize the prediction error and obtain the trained equipment aging rate prediction model.

[0041] Preferably, the prediction unit is used to input the feature vector G into the trained equipment aging rate prediction model to obtain the equipment aging rate v of each electrical equipment. aming , and according to the equipment aging rate v aming Get the remaining service life RUL of the equipment. Taking electrical equipment k as an example, the specific method for obtaining the remaining service life RUL(k) of electrical equipment k is:

[0042]

[0043] Where, t used (k) represents the cumulative usage time of electrical equipment k at the current time point t, v aming (t, k) represents the aging rate of electrical equipment k at the current time point, Indicates the total life of electrical equipment k;

[0044] Based on the eigenvector G and the equipment's remaining service life RUL, a weighted fusion calculation is performed to obtain the equipment's comprehensive fault score Gz. The specific method for obtaining the equipment's comprehensive fault score Gz is as follows:

[0045]

[0046] Where α1, α2, α3 and α4 represent the remaining service life RUL of the equipment and the temperature-discharge intensity correlation factor Xg, respectively. p，T , the weight coefficient of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), and C represents the first correction constant.

[0047] Preferably, the intelligent early warning feedback module is used to preset a first device failure threshold λ1 and a second device failure threshold λ2, and for each abnormal electrical device in the abnormal device group, compare and analyze the device comprehensive failure score Gz of the abnormal electrical device with the first device failure threshold λ1 and the second device failure threshold λ2 to evaluate the failure risk level of the electrical device. The specific evaluation content is as follows:

[0048] If the comprehensive equipment fault score Gz is less than or equal to the first equipment fault threshold λ1, the fault risk level of the electrical equipment is determined to be the first risk level, and the electrical equipment of the first risk level is marked as green. Regular inspection and monitoring are carried out, and the operating data of the electrical equipment is collected;

[0049] If the comprehensive equipment failure score Gz is greater than the first equipment failure threshold λ1 and less than the second equipment failure threshold λ2, the electrical equipment is determined to have a second risk level and marked yellow. The inspection frequency of the electrical equipment is then changed from semi-annual to quarterly, and preventive maintenance measures are implemented, including cleaning dust inside the equipment and replacing worn parts.

[0050] If the comprehensive equipment failure score Gz is greater than or equal to the second equipment failure threshold λ2, the fault risk level of the electrical equipment is determined to be the third risk level. At this time, a danger alarm is immediately triggered, the electrical equipment of the third risk level is marked red, and the operation of the electrical equipment is immediately stopped. At the same time, a danger alarm is sent to the person in charge of the electrical equipment until the person in charge of the electrical equipment responds. For all electrical equipment marked red, a failure mode impact analysis is performed, and emergency repairs are carried out. The power-related data after the repairs are continuously monitored;

[0051] The intelligent feedback unit is used to collect power-related data after maintenance and feed the power-related data back to the aging model prediction module to dynamically update the equipment aging rate prediction model.

[0052] Preferably, the electrical power equipment fault monitoring method based on the industrial Internet includes the following steps:

[0053] Step 1: Collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center;

[0054] Step 2: Based on the power-related data set H, calculate and obtain the discharge intensity change rate Δp, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group;

[0055] Step 3: extract characteristic indicators of each electrical device in the abnormal electrical device group and construct a characteristic vector G;

[0056] Step 4: Build a prediction model for equipment aging rate and obtain the equipment aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment;

[0057] Step 5: Based on the equipment comprehensive fault score Gz, conduct a risk assessment on the electrical equipment, classify the fault risk level of the electrical equipment, and trigger the intelligent early warning mechanism.

[0058] The present invention provides an electrical power equipment fault monitoring method and system based on the Industrial Internet, which has the following beneficial effects:

[0059] (1) By integrating intelligent sensor groups, various data of equipment operation are collected in real time, such as equipment usage time t used , electrical equipment temperature T, electrical equipment total current I total , voltage V, insulation resistance r ins and partial discharge intensity p discharge etc., and upload power-related data to the local server for pre-processing through wireless technology. This data collection method ensures real-time monitoring of the equipment's operating status and rapid data updates, allowing the health status of the power equipment to be captured and fed back to the data center in a timely manner. By calculating and comparing the discharge intensity change rate Δp, the system can automatically trigger an early warning mechanism and quickly identify abnormal equipment conditions. This real-time monitoring and rapid early warning capability improves the response speed of fault detection, avoids delays and omissions caused by manual inspections in traditional monitoring methods, and reduces the risk of equipment downtime due to faults.

[0060] (2) Through the feature extraction module, based on the real-time collected power-related data, a number of key feature indicators are extracted, such as the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan (δ), and input these characteristic indicators into the equipment aging rate prediction model, and use the linear regression algorithm to predict the equipment aging rate v of electrical equipment. aming Combined with the prediction results, the system can accurately calculate the remaining useful life (RUL) of the equipment. This data enables equipment managers to understand the aging process of the equipment in real time, make scientific maintenance decisions, and avoid sudden equipment failures that affect system stability. In addition, the generation of the equipment comprehensive fault score Gz also provides a quantitative assessment of equipment health, which helps to optimize maintenance cycles and arrange appropriate preventive maintenance, extend equipment service life and reduce unplanned downtime.

[0061] (3) The equipment failure risk level is evaluated through the equipment comprehensive fault score Gz, and the risk is automatically classified according to the preset fault threshold. By dividing the equipment into different risk levels, the system can provide personalized maintenance strategies according to the risk level of the equipment. For example, electrical equipment of the first risk level will be inspected regularly to maintain its good condition. Electrical equipment of the second risk level will increase the inspection frequency and implement preventive maintenance. Electrical equipment of the third risk level will immediately trigger an emergency maintenance warning, stop the equipment operation, and prevent major failures. This risk classification and intelligent feedback mechanism ensures that each electrical equipment can be dealt with in a targeted and timely manner, which can improve the management efficiency of the power system and reduce the impact of equipment failures on the entire power grid. At the same time, the real-time feedback and adjustment mechanism also promotes the dynamic update of the aging model prediction module, continuously optimizes equipment health prediction and risk assessment, and provides intelligent support for power equipment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 This is a block diagram of the electrical power equipment fault monitoring system based on the Industrial Internet of the present invention;

[0063] Figure 2 This is a flow chart of a method for monitoring faults of electrical power equipment based on the Industrial Internet according to the present invention;

[0064] Figure 3 This is a schematic diagram of the calculation process of the comprehensive fault score Gz of the equipment of the present invention;

[0065] Figure 4 This is a schematic diagram of the risk level assessment of electrical equipment according to the present invention. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0067] Example 1

[0068] See also Figure 1 , the present invention provides an electrical power equipment fault monitoring system based on the industrial Internet, including a data acquisition and preprocessing module, a data analysis module, a feature extraction module, an aging model prediction module and an intelligent early warning feedback module;

[0069] The data acquisition and preprocessing module is used to collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center;

[0070] The data analysis module is used to calculate the discharge intensity change rate Δp based on the power-related data set H, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group;

[0071] The feature extraction module is used to extract the characteristic indicators of each electrical device in the abnormal electrical equipment group and construct a feature vector G;

[0072] The aging model prediction module is used to build a device aging rate prediction model and obtain the device aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment;

[0073] The intelligent early warning feedback module is used to conduct risk assessment on electrical equipment based on the equipment comprehensive fault score Gz, classify the fault risk level of electrical equipment, and trigger the intelligent early warning mechanism.

[0074] In the embodiment, through the real-time data collection and preprocessing of the intelligent sensor group, all-round monitoring of power equipment is achieved, and the timeliness and accuracy of equipment fault detection are improved. The data collection and preprocessing module accurately obtains the power-related data of the electrical equipment and uploads it to the power data center after preprocessing, ensuring the high quality and real-time nature of the data. The data analysis module can quickly identify abnormal changes in the equipment status by calculating the discharge intensity change rate Δp, and automatically trigger the early warning mechanism to timely screen out electrical equipment that may fail and build an abnormal equipment group. The feature extraction module further extracts the key features of the abnormal electrical equipment and generates a feature vector G, providing accurate data support for subsequent fault prediction. The aging model prediction module accurately predicts the aging rate v of the equipment through a linear regression algorithm. aming Combined with the equipment's remaining useful life (RUL), the eigenvector G is used to calculate the equipment's comprehensive fault score (Gz), providing a quantitative basis for fault risk assessment. Finally, the intelligent early warning feedback module performs risk assessment and grading based on the equipment's comprehensive fault score, ensuring a timely response from the fault warning mechanism and preventing sudden equipment failures. This system not only improves the accuracy of equipment failure prediction but also optimizes maintenance decisions, enhancing the operational safety and reliability of power equipment and reducing equipment downtime and repair costs.

[0075] Example 2

[0076] Please refer to Figure 1 and Figure 3,Specifically: the data acquisition and preprocessing module includes a data acquisition unit and a data ,processing unit;

[0077] The data acquisition unit is used to deploy smart sensor groups on electrical equipment and use the smart sensor groups to collect power-related data generated during the operation of the power equipment in real time, where the power-related data includes the cumulative usage time t used , electrical equipment temperature T, electrical equipment total current I total , voltage V, insulation resistance r ins and partial discharge intensity p discharge ;

[0078] Equipment usage time t used Obtained through electrical equipment data recording systems;

[0079] The temperature T of the electrical equipment is obtained by using a temperature sensor;

[0080] Total current of electrical equipment I total Obtained by using a current monitoring sensor;

[0081] The voltage V is obtained by using a voltage sensor;

[0082] Insulation resistance r ins Obtained by using an insulation resistance sensor;

[0083] Partial discharge intensity p discharge Acquired by using partial discharge sensors;

[0084] The intelligent sensor group includes temperature sensor, current monitoring sensor, voltage sensor, insulation resistance sensor and partial discharge sensor;

[0085] The real-time collected power-related data is transmitted to the local server via wireless Wi-Fi technology.

[0086] The data processing unit is used to process the power-related data in the local server, including anomaly detection and correction, denoising and missing value filling;

[0087] Anomaly detection and correction refers to using the isolation forest algorithm to detect abnormal data points in the data and correct the outliers based on the neighboring data;

[0088] Isolation Forest Algorithm is a tree-based machine learning algorithm used for anomaly detection. Its basic principle is to "isolate" data points by randomly selecting features and randomly selecting split points, and then determine which data points are abnormal.

[0089] Denoising refers to using wavelet transform denoising method to decompose the noise in the data into different frequency components and remove the high-frequency noise components;

[0090] Missing value filling refers to filling missing values in the time series data of equipment operating status using spline interpolation. Spline interpolation can take into account the continuity of time and maintain data consistency on the time axis. It is particularly effective when the continuity of electrical equipment parameters, such as current and voltage, is strong.

[0091] Based on the processed power-related data, a power-related data set H is constructed and stored in a power data center, wherein the power-related data set H includes power-related data generated by several electrical devices during operation.

[0092] In the embodiment, by deploying an intelligent sensor group, power-related data of electrical equipment during operation is collected in real time, providing comprehensive data support for the health status of the equipment, and transmitting data via wireless Wi-Fi, ensuring the real-time and remote accessibility of the data, so that the operation status of the equipment can be quickly fed back to the local server for processing. The data processing unit uses an advanced isolation forest algorithm for anomaly detection and correction, which can accurately identify and correct anomalies in the data to ensure the accuracy of the data. At the same time, the wavelet transform denoising method can effectively remove high-frequency noise and maintain the original signal of the data, which helps to avoid noise interference and improve the reliability of data analysis. In addition, the spline interpolation method is used to fill in missing values. Taking into account the continuity and consistency of time series data, it can accurately restore missing data in the operation of the equipment and ensure the integrity of the equipment status data. Through data processing, the system can construct a high-quality power-related data set H, which provides accurate and reliable basic data support for subsequent data analysis and fault prediction, thereby improving the intelligence and accuracy of the entire fault monitoring system.

[0093] Example 3

[0094] Please refer to Figure 1 and Figure 3 Specifically, the data analysis module is used to analyze the operating status of electrical equipment over a period of time based on the power-related data set H, and calculate the discharge intensity change rate Δp. The specific calculation process of the discharge intensity change rate Δp is:

[0095]

[0096] Where p discharge (t) represents the partial discharge intensity of the electrical equipment at the current time point t, p discharge (t-1) represents the partial discharge intensity of the electrical equipment at the previous time point t-1;

[0097] The discharge intensity change rate Δp refers to the degree of change in the partial discharge intensity inside electrical equipment within a certain time interval. The partial discharge intensity is an important signal of the aging of the insulation material of electrical equipment. An increase in partial discharge intensity usually indicates the degradation of the equipment's insulation layer and may be a sign of impending equipment failure.

[0098] A partial discharge change threshold Fd is preset, and the discharge intensity change rate Δp is compared with the partial discharge change threshold Fd. When the discharge intensity change rate Δp is greater than or equal to the partial discharge change threshold Fd, that is, Δp ≥ Fd, the early warning mechanism is automatically triggered. At this time, the electrical equipment is judged to be in an abnormal state, the electrical equipment in the abnormal state is marked, and an abnormal electrical equipment group is constructed.

[0099] The following are some examples:

[0100] Assume that there is an electrical device, the partial discharge intensity collected at time point t1 is 50pC, the partial discharge intensity collected at time point t2 is 80pC, and the time interval between time point t1 and time point t2 is 10 minutes, that is, p discharge (t2)=80pC,p discharge (t1) = 50 pC;

[0101] Calculate the rate of change of discharge intensity:

[0102]

[0103] The preset partial discharge change threshold Fd is 2 pC / min. At this time, if the discharge intensity change rate Δp=3 pC / min>2 pC / min, it is determined that the operating state of the electrical equipment is abnormal.

[0104] In an embodiment, by analyzing power-related data and calculating the discharge intensity change rate Δp, abnormal changes in the equipment operating status can be monitored in real time. When the local discharge intensity of the equipment changes, by comparing it with the preset discharge change threshold Fd, the system can promptly identify potential fault hazards of the equipment. If the discharge intensity change rate Δp is greater than or equal to the local discharge change threshold Fd, the system will automatically trigger the early warning mechanism and determine that the equipment operating status is abnormal. This process can not only quickly discover potential problems in equipment operation, but also mark and classify equipment showing signs of failure through the construction of abnormal electrical equipment groups to ensure timely processing. In this way, equipment failures can be warned in advance, downtime and maintenance costs caused by equipment failures can be reduced, the maintenance strategy of power equipment can be optimized, and the stability and reliability of the power system can be improved. This step effectively improves the response speed and accuracy of the system, making electrical equipment management more intelligent and efficient.

[0105] Example 4

[0106] Please refer to Figure 1 and Figure 3 Specifically: the feature extraction module is used to obtain the power-related data of each abnormal electrical device in the abnormal electrical device group based on the power-related data set H, and extract the characteristic indicators of each abnormal electrical device, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ);

[0107] Analyze the relationship between temperature and discharge intensity to obtain the temperature-discharge intensity correlation factor Xg p，T , where the temperature-discharge intensity correlation factor Xg p，T The specific way to obtain it is:

[0108]

[0109] In the formula, n represents the total number of data sampling times, p discharge (t i ) represents the time point t i The partial discharge intensity at represents the mean value of the local discharge intensity, T(t i ) represents the time point t i The temperature of the electrical equipment at represents the average temperature of the electrical equipment, i represents the data sampling index, i = {1, 2, 3, ..., n}.

[0110] Temperature-discharge intensity correlation factor Xg p，T It is an indicator obtained by analyzing the relationship between the temperature change of electrical equipment and the change of partial discharge intensity. It reflects whether the temperature change of electrical equipment during operation has a certain degree of dependence on the change of partial discharge intensity. In practical applications, discharge intensity is used to reflect the local aging behavior of equipment insulation materials, and temperature is affected by environmental changes, equipment load and faults. Discharge intensity and temperature both have continuous time characteristics. Therefore, the temperature-discharge intensity correlation factor Xg p，T It is a characteristic indicator that evolves over time.

[0111] Analyze the insulation resistance fluctuation amplitude between consecutive time points to obtain the insulation resistance fluctuation coefficient JY. The insulation resistance fluctuation coefficient JY is obtained as follows:

[0112]

[0113] Where r ins (t i ) represents the time point t i The insulation resistance of electrical equipment, rins (t i-1 ) Time point t i-1 Insulation resistance of electrical equipment;

[0114] In electrical equipment, insulation resistance is an important indicator for determining whether the equipment can operate normally. As the equipment is used for a longer time, the insulation material will gradually age, causing the insulation resistance to decrease. The insulation resistance fluctuation coefficient JY, as a characteristic indicator, can measure the amplitude of the change in insulation resistance and its stability during operation, thereby more comprehensively reflecting the health status of the equipment.

[0115] Analyze the power loss of the insulation material of the electrical equipment to calculate the dielectric loss factor tan(δ). The dielectric loss factor tan(δ) is obtained as follows:

[0116]

[0117] Where V represents voltage, I total It represents the total current of the electrical equipment, cos(φ) represents the phase difference between the voltage and current waveforms, which can be obtained by using an oscilloscope to identify, R reactive Represents the reaction resistance, which is obtained from the electrical equipment resistance material database;

[0118] The dielectric loss factor tan(δ) is a physical quantity that describes the energy loss of insulating materials in electrical equipment under the action of an electric field. It is used to reflect the degree of energy loss of insulating materials under AC voltage. The larger the dielectric loss factor tan(δ), the higher the electrical energy loss of the insulating material. By monitoring the dielectric loss factor tan(δ), early warning of electrical equipment failures and health status assessment can be achieved, avoiding equipment failures due to insulation problems, thereby improving the safety and reliability of electrical equipment.

[0119] Based on the extracted characteristic indicators, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ), construct a characteristic vector G, where the characteristic vector G includes characteristic indicators of each electrical device.

[0120] In the embodiment, the temperature-discharge intensity correlation factor Xg is calculated by extracting the characteristics of each electrical device in the abnormal electrical device group. p，TThe system can analyze the relationship between temperature changes and partial discharge intensity of electrical equipment, helping to reveal possible abnormal behavior and early signs of aging during equipment operation. By calculating the insulation resistance fluctuation coefficient JY, the stability of the equipment's insulation material can be quantified, its fluctuation amplitude under different operating conditions can be detected, and the risk of insulation degradation can be detected early. In addition, the calculation of the dielectric loss factor tan(δ) can reflect the energy loss of the insulation material of electrical equipment. By combining the voltage, current, and the phase difference between the voltage and current waveforms, the system can more accurately assess the equipment's energy loss and health status. The extraction of these characteristic indicators not only provides strong data support for subsequent aging models, but also helps to monitor the aging process of the equipment in real time and improve the accuracy of fault prediction. Ultimately, the constructed characteristic vector G integrates these key indicators, providing the system with a comprehensive and accurate equipment health assessment foundation, effectively improving equipment risk assessment and early warning capabilities, and reducing the probability of failure.

[0121] Example 5

[0122] Please refer to Figure 1 and Figure 3 ,Specifically: the aging model prediction module includes a model building unit and a ,prediction unit;

[0123] The model building unit is used to build an equipment aging rate prediction model using a linear regression algorithm and obtain historical power-related data of each abnormal electrical equipment in the power data center. The specific form of the equipment aging rate prediction model is:

[0124] v aming =ω1·Xg p，T +ω2·JY+ω3·tan(δ)+b;

[0125] Where ω1, ω2 and ω3 represent the temperature-discharge intensity correlation factor Xg p，T , the regression coefficients of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), b represents the bias term, where the regression coefficients ω1, ω2 and ω3 are obtained by using the minimum mean square error MSE method;

[0126] Feature extraction is performed on the historical power-related data of each abnormal electrical equipment, and the extracted feature indicators are input into the equipment aging rate prediction model for model training. The mean square error (MSE) is used as the loss function to minimize the prediction error and obtain the trained equipment aging rate prediction model.

[0127] The linear regression algorithm is a machine learning algorithm used to establish a linear relationship between input features and output results. The equipment aging rate prediction model constructed using the linear regression algorithm can find the relationship between each characteristic indicator and the equipment aging rate based on the collected historical power-related data, and predict the future aging of the equipment.

[0128] The prediction unit is used to input the feature vector G into the trained equipment aging rate prediction model to obtain the equipment aging rate v of each electrical equipment aming , and according to the equipment aging rate v aming Get the remaining service life RUL of the equipment. Taking electrical equipment k as an example, the specific method for obtaining the remaining service life RUL(k) of electrical equipment k is:

[0129]

[0130] Where, t used (k) represents the cumulative usage time of electrical equipment k at the current time point t, v aming (t, k) represents the aging rate of electrical equipment k at the current time point, Indicates the total life of electrical equipment k. In actual applications, the status of electrical equipment is affected by various complex factors, and its life can easily be "prematurely consumed" or "delayed". Therefore, it is necessary to consider the influence of various factors and dynamically analyze the remaining service life (RUL) of electrical equipment;

[0131] Based on the eigenvector G and the equipment's remaining service life RUL, a weighted fusion calculation is performed to obtain the equipment's comprehensive fault score Gz. The specific method for obtaining the equipment's comprehensive fault score Gz is as follows:

[0132]

[0133] Where α1, α2, α3 and α4 represent the remaining service life RUL of the equipment and the temperature-discharge intensity correlation factor Xg, respectively. p，T , the weight coefficients of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), C represents the first correction constant, wherein the values of the weight coefficients α1, α2, α3 and α4 are set by experts according to actual conditions, wherein the change of the remaining service life RUL of the equipment is negatively correlated with the comprehensive fault score Gz of the equipment, that is, the higher the remaining service life RUL of the equipment, the lower the comprehensive fault score Gz of the equipment, therefore the weight coefficient α1 of the remaining service life RUL of the equipment is preceded by a negative sign.

[0134] In the embodiment, a device aging rate prediction model is constructed by a linear regression algorithm, which can accurately predict the aging rate of the device. The device aging rate prediction model is trained using historical power data and optimizes the prediction results by minimizing the mean square error MSE loss function to ensure the high accuracy of the prediction model. The prediction unit calculates the device aging rate v according to the real-time feature vector G of the device input according to the trained device aging rate prediction model. aming , and further calculate the remaining service life RUL of the equipment. By combining the remaining service life RU of the equipment with the characteristic vector G, the comprehensive fault score Gz of the equipment is obtained, which is used to evaluate the health status of the equipment. This prediction model and scoring mechanism not only provides a scientific basis for equipment management, but also realizes real-time evaluation and optimization of the health status of the equipment, avoids excessive maintenance or delayed repairs, reduces maintenance costs and improves the operational reliability of the equipment. This process greatly improves the intelligent level of equipment management, enhances fault warning and decision support capabilities, and provides a guarantee for the stable operation of the power system.

[0135] Example 6

[0136] Please refer to Figure 1 、 Figure 3 and Figure 4 Specifically, the intelligent early warning feedback module is used to preset a first device fault threshold λ1 and a second device fault threshold λ2, and for each abnormal electrical device in the abnormal device group, compare and analyze the device comprehensive fault score Gz of the abnormal electrical device with the first device fault threshold λ1 and the second device fault threshold λ2 to evaluate the fault risk level of the electrical device. The specific evaluation contents are as follows:

[0137] If the comprehensive equipment fault score Gz is less than or equal to the first equipment fault threshold λ1, the fault risk level of the electrical equipment is determined to be the first risk level, and the electrical equipment of the first risk level is marked as green. Regular inspection and monitoring are carried out, and the operating data of the electrical equipment is collected;

[0138] If the comprehensive equipment failure score Gz is greater than the first equipment failure threshold λ1 and less than the second equipment failure threshold λ2, the electrical equipment is determined to have a second risk level and marked yellow. The inspection frequency of the electrical equipment is then changed from semi-annual to quarterly, and preventive maintenance measures are implemented, including cleaning dust inside the equipment and replacing worn parts.

[0139] If the comprehensive equipment failure score Gz is greater than or equal to the second equipment failure threshold λ2, the fault risk level of the electrical equipment is determined to be the third risk level. At this time, a danger alarm is immediately triggered, the electrical equipment of the third risk level is marked red, and the operation of the electrical equipment is immediately stopped. At the same time, a danger alarm is sent to the person in charge of the electrical equipment until the person in charge of the electrical equipment responds. For all electrical equipment marked red, a failure mode impact analysis is performed, and emergency repairs are carried out. The power-related data after the repairs are continuously monitored;

[0140] The intelligent feedback unit is used to collect power-related data after maintenance and feed the power-related data back to the aging model prediction module to dynamically update the equipment aging rate prediction model.

[0141] In the embodiment, by combining the fault warning unit and the intelligent feedback unit, accurate assessment and timely response to the failure risk of electrical equipment are achieved. First, the fault warning unit further divides the risk level of electrical equipment according to the equipment comprehensive fault score Gz and the preset fault thresholds λ1 and λ2. This grading mechanism ensures that equipment with different risk levels can be treated differently. Electrical equipment marked as green requires regular monitoring, electrical equipment marked as yellow requires increased inspection frequency and preventive maintenance, and electrical equipment marked as red needs to be stopped immediately, triggering emergency maintenance and failure mode impact analysis, thereby avoiding potential major failures. This intelligent, hierarchical management model improves the accuracy and response speed of equipment fault warnings, effectively reducing the impact of electrical equipment failures on the power system. The intelligent feedback unit collects data after maintenance and feeds it back to the aging model prediction module, realizing dynamic updates and model optimization, further improving the accuracy and timeliness of equipment health prediction. In short, this module realizes the early identification, accurate processing and continuous optimization of electrical equipment failure risks, provides intelligent decision support for equipment maintenance, reduces downtime, and improves the stability and reliability of the system.

[0142] Example 7

[0143] Please refer to Figure 2 ,Specifically: The electrical power equipment fault monitoring method based on industrial Internet, includes the following steps,

[0144] Step 1: Collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center;

[0145] Step 2: Based on the power-related data set H, calculate and obtain the discharge intensity change rate Δp, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group;

[0146] Step 3: extract characteristic indicators of each electrical device in the abnormal electrical device group and construct a characteristic vector G;

[0147] Step 4: Build a prediction model for equipment aging rate and obtain the equipment aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment;

[0148] Step 5: Based on the equipment comprehensive fault score Gz, conduct a risk assessment on the electrical equipment, classify the fault risk level of the electrical equipment, and trigger the intelligent early warning mechanism.

[0149] In the embodiment, the fault prevention capability and management efficiency of electrical equipment are improved through real-time data collection, intelligent analysis and prediction mechanisms. First, power-related data is collected in real time, and through data preprocessing and uploading, the monitoring of equipment status is made more real-time and accurate. The power data set H is extracted, the discharge intensity change rate Δp is calculated, the early warning mechanism is triggered, equipment anomalies are quickly identified, and an abnormal equipment group is constructed. Then, through the feature extraction module, the system can generate a feature vector G based on the key indicators of the electrical equipment, and combine it with the linear regression algorithm to accurately predict the equipment aging rate v aming , and calculate the remaining service life RUL of the electrical equipment. Finally, the comprehensive equipment failure score Gz is constructed to provide a basis for risk assessment. The failure risk level is divided according to the health status of the equipment, and the intelligent early warning mechanism is activated to ensure that the equipment is maintained and processed in a timely manner, optimize the equipment maintenance decision, reduce the risk of failure, and improve the stability and safety of the power system.

[0150] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. The electrical power equipment fault monitoring system based on the Industrial Internet is characterized by: It includes data acquisition and preprocessing module, data analysis module, feature extraction module, aging model prediction module and intelligent early warning feedback module; The data acquisition and preprocessing module is used to collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center; The data analysis module is used to calculate the discharge intensity change rate Δp based on the power-related data set H, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group; The feature extraction module is used to extract the characteristic indicators of each electrical device in the abnormal electrical equipment group and construct a feature vector G; The aging model prediction module is used to build a device aging rate prediction model and obtain the device aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment; The intelligent early warning feedback module is used to conduct risk assessment on electrical equipment based on the equipment comprehensive fault score Gz, classify the fault risk level of electrical equipment, and trigger the intelligent early warning mechanism.

2. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 1 is characterized in that: The data acquisition and preprocessing module includes a data acquisition unit and a data processing unit; The data acquisition unit is used to deploy smart sensor groups on electrical equipment and use the smart sensor groups to collect power-related data generated during the operation of the power equipment in real time, where the power-related data includes the cumulative usage time t used , electrical equipment temperature T, electrical equipment total current I total , voltage V, insulation resistance r ins and partial discharge intensity p discharge ; The intelligent sensor group includes temperature sensor, current monitoring sensor, voltage sensor, insulation resistance sensor and partial discharge sensor; The real-time collected power-related data is transmitted to the local server via wireless Wi-Fi technology.

3. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 2 is characterized in that: The data processing unit is used to process the power-related data in the local server, including anomaly detection and correction, denoising and missing value filling; Anomaly detection and correction refers to using the isolation forest algorithm to detect abnormal data points in the data and correct the outliers based on the neighboring data; Denoising refers to using wavelet transform denoising method to decompose the noise in the data into different frequency components and remove the high-frequency noise components; Missing value filling refers to filling missing values in the time series data of equipment operation status using spline interpolation method; Based on the processed power-related data, a power-related data set H is constructed and stored in a power data center, wherein the power-related data set H includes power-related data generated by several electrical devices during operation.

4. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 3 is characterized in that: The data analysis module is used to analyze the operating status of electrical equipment over a period of time based on the power-related data set H and calculate the discharge intensity change rate Δp. The specific calculation process of the discharge intensity change rate Δp is: Where p discharge (t) represents the partial discharge intensity of the electrical equipment at the current time point t, p discharge (t-1) represents the partial discharge intensity of the electrical equipment at the previous time point t-1; A partial discharge change threshold Fd is preset, and the discharge intensity change rate Δp is compared with the partial discharge change threshold Fd. When the discharge intensity change rate Δp is greater than or equal to the partial discharge change threshold Fd, that is, Δp ≥ Fd, the early warning mechanism is automatically triggered. At this time, the electrical equipment is judged to be in an abnormal state, the electrical equipment in the abnormal state is marked, and an abnormal electrical equipment group is constructed.

5. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 4 is characterized in that: The feature extraction module is used to obtain the power-related data of each abnormal electrical device in the abnormal electrical equipment group based on the power-related data set H, and extract the characteristic indicators of each abnormal electrical device, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ); Analyze the relationship between temperature and discharge intensity to obtain the temperature-discharge intensity correlation factor Xg p，T , where the temperature-discharge intensity correlation factor Xg p，T The specific way to obtain it is: In the formula, n represents the total number of data sampling times, p discharge (t i ) represents the time point t i The partial discharge intensity at represents the mean value of the local discharge intensity, T(t i ) represents the time point t i The temperature of the electrical equipment at represents the average temperature of the electrical equipment, i represents the data sampling index, i = {1, 2, 3, ..., n}.

6. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 5 is characterized in that: Analyze the insulation resistance fluctuation amplitude between consecutive time points to obtain the insulation resistance fluctuation coefficient JY. The insulation resistance fluctuation coefficient JY is obtained as follows: Where r ins (t i ) represents the time point t i The insulation resistance of electrical equipment, r ins (t i-1 ) Time point t i-1 Insulation resistance of electrical equipment; Analyze the power loss of the insulation material of the electrical equipment to obtain the dielectric loss factor tan(δ). The dielectric loss factor tan(δ) is obtained as follows: Where V represents voltage, I total It represents the total current of the electrical equipment, cos(φ) represents the phase difference between the voltage and current waveforms, R reactive represents the reaction resistance; Based on the extracted characteristic indicators, including the temperature-discharge intensity correlation factor Xg p，T , insulation resistance fluctuation coefficient JY and dielectric loss factor tan(δ), construct a characteristic vector G, where the characteristic vector G includes characteristic indicators of each electrical device.

7. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 1 is characterized in that: The aging model prediction module includes a model building unit and a prediction unit; The model building unit is used to build an equipment aging rate prediction model using a linear regression algorithm and obtain historical power-related data of each abnormal electrical equipment in the power data center. The specific form of the equipment aging rate prediction model is: v aming =ω1·Xg p，T +ω2·JY+ω3·tan(δ)+b; Where ω1, ω2 and ω3 represent the temperature-discharge intensity correlation factor Xg p，T , the regression coefficients of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), b represents the bias term; Feature extraction is performed on the historical power-related data of each abnormal electrical equipment, and the extracted feature indicators are input into the equipment aging rate prediction model for model training. The mean square error (MSE) is used as the loss function to minimize the prediction error and obtain the trained equipment aging rate prediction model.

8. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 7 is characterized in that: The prediction unit is used to input the feature vector G into the trained equipment aging rate prediction model to obtain the equipment aging rate v of each electrical equipment aming , and according to the equipment aging rate v aming Get the remaining service life RUL of the equipment. Taking electrical equipment k as an example, the specific method for obtaining the remaining service life RUL(k) of electrical equipment k is: Where, t used (k) represents the cumulative usage time of electrical equipment k at the current time point t, v aming (t, k) represents the aging rate of electrical equipment k at the current time point, Indicates the total life of electrical equipment k; Based on the eigenvector G and the equipment's remaining service life RUL, a weighted fusion calculation is performed to obtain the equipment's comprehensive fault score Gz. The specific method for obtaining the equipment's comprehensive fault score Gz is as follows: Where α1, α2, α3 and α4 represent the remaining service life RUL of the equipment and the temperature-discharge intensity correlation factor Xg, respectively. p，T , the weight coefficient of the insulation resistance fluctuation coefficient JY and the dielectric loss factor tan(δ), and C represents the first correction constant.

9. The electrical power equipment fault monitoring system based on the Industrial Internet according to claim 8 is characterized in that: The intelligent early warning feedback module is used to preset a first device failure threshold λ1 and a second device failure threshold λ2. For each abnormal electrical device in the abnormal device group, the module compares and analyzes the device comprehensive failure score Gz of the abnormal electrical device with the first device failure threshold λ1 and the second device failure threshold λ2 to assess the failure risk level of the electrical device. The specific assessment contents are as follows: If the comprehensive equipment fault score Gz is less than or equal to the first equipment fault threshold λ1, the fault risk level of the electrical equipment is determined to be the first risk level, and the electrical equipment of the first risk level is marked as green. Regular inspection and monitoring are carried out, and the operating data of the electrical equipment is collected; If the comprehensive equipment failure score Gz is greater than the first equipment failure threshold λ1 and less than the second equipment failure threshold λ2, the electrical equipment is determined to have a second risk level and marked yellow. The inspection frequency of the electrical equipment is then changed from semi-annual to quarterly, and preventive maintenance measures are implemented, including cleaning dust inside the equipment and replacing worn parts. If the comprehensive equipment failure score Gz is greater than or equal to the second equipment failure threshold λ2, the fault risk level of the electrical equipment is determined to be the third risk level. At this time, a danger alarm is immediately triggered, the electrical equipment of the third risk level is marked red, and the operation of the electrical equipment is immediately stopped. At the same time, a danger alarm is sent to the person in charge of the electrical equipment until the person in charge of the electrical equipment responds. For all electrical equipment marked red, a failure mode impact analysis is performed, and emergency repairs are carried out. The power-related data after the repairs are continuously monitored; The intelligent feedback unit is used to collect power-related data after maintenance and feed the power-related data back to the aging model prediction module to dynamically update the equipment aging rate prediction model.

10. An electrical power equipment fault monitoring method based on the industrial Internet, used to implement the electrical power equipment fault monitoring system based on the industrial Internet as described in any one of claims 1 to 9, characterized in that: The following steps are included: Step 1: Collect power-related data during the operation of electrical equipment, construct a power-related data set H, perform preprocessing, and upload the preprocessed power-related data set H to the power data center; Step 2: Based on the power-related data set H, calculate and obtain the discharge intensity change rate Δp, and compare and analyze the discharge intensity change rate Δp with the preset partial discharge change threshold Fd to screen and construct an abnormal electrical equipment group; Step 3: extract characteristic indicators of each electrical device in the abnormal electrical device group and construct a characteristic vector G; Step 4: Build a prediction model for equipment aging rate and obtain the equipment aging rate v aming , and calculate the remaining service life RUL of the equipment, combined with the characteristic vector G, to obtain the comprehensive fault score Gz of the equipment; Step 5: Based on the equipment comprehensive fault score Gz, conduct a risk assessment on the electrical equipment, classify the fault risk level of the electrical equipment, and trigger the intelligent early warning mechanism.

Images