Substation partial discharge fault early warning method and system based on artificial intelligence

Abstract

The invention discloses a substation partial discharge fault early warning method and system based on artificial intelligence, and relates to the technical field of substation partial discharge fault early warning. Performing feature extraction on the substation partial discharge abnormal data to obtain an electrical parameter abnormal coefficient, an equipment operation abnormal coefficient and an environment abnormal coefficient; constructing a transformer substation partial discharge anomaly model based on machine learning by using the electrical parameter anomaly coefficient, the equipment operation anomaly coefficient and the environment anomaly coefficient, and generating a transformer substation partial discharge anomaly coefficient according to the transformer substation partial discharge anomaly model; and carrying out comparative analysis on the substation partial discharge abnormal coefficient based on a preset historical substation partial discharge stability coefficient, judging whether the substation partial discharge condition has a fault or not, and carrying out early warning processing. According to the invention, the transformer substation partial discharge abnormity is analyzed through artificial intelligence, and the problem that the transformer substation partial discharge fault cannot be timely and effectively warned is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of partial discharge fault early warning for substations, and more specifically, to the technical field of partial discharge fault early warning method and system for substations based on artificial intelligence. Background Art

[0002] Partial discharge in substations is a common phenomenon in high-voltage electrical equipment. It refers specifically to non-penetrating discharges that occur in local areas inside or on the insulating medium of electrical equipment due to excessive electric field strength or insulation material defects. This phenomenon mainly occurs in high-voltage equipment such as transformers, mutual inductors, and GIS, and has an important impact on the safe operation of substations.

[0003] In substations, partial discharge often indicates potential problems or defects in the insulation system. Long-term partial discharge will accelerate the aging and damage of insulation materials, and may even cause more serious electrical faults such as insulation breakdown or flashing.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a substation partial discharge fault early warning method and system based on artificial intelligence, which solves the problem that substation partial discharge faults cannot be warned in a timely and effective manner through artificial intelligence analysis of substation partial discharge anomalies.

[0006] To achieve the above object, the present invention provides the following technical solutions: The artificial intelligence-based substation partial discharge fault early warning method and system are characterized by comprising the following steps: obtaining substation partial discharge abnormal data, performing feature extraction on the substation partial discharge abnormal data to obtain electrical parameter abnormal coefficients, equipment operation abnormal coefficients and environmental abnormal coefficients; constructing a substation partial discharge abnormal model based on machine learning for the electrical parameter abnormal coefficients, equipment operation abnormal coefficients and environmental abnormal coefficients, and generating a substation partial discharge abnormal coefficient based on the substation partial discharge abnormal model; performing comparative analysis on the substation partial discharge abnormal coefficient based on a preset historical substation partial discharge stability coefficient, judging whether the substation partial discharge situation has a fault and performing early warning processing.

[0007] In a preferred embodiment, the electrical parameter anomaly coefficient is specifically obtained by the following method: obtaining the voltage and current waveforms of the substation equipment, the actual capacity of the transformer and the transient waveform data; obtaining the harmonic order and harmonic amplitude of the obtained voltage and current waveforms of the substation equipment based on the fast Fourier transform algorithm; dividing the substation working time period G into N nodes, and monitoring and obtaining the grid frequency fluctuation data of each node in the G time; obtaining the grid frequency fluctuation deviation value of each node based on the preset normal grid frequency fluctuation data according to the grid frequency fluctuation data; performing data processing on the grid frequency fluctuation deviation value of each node to obtain the grid frequency fluctuation anomaly coefficient; calculating the transformer load rate based on the preset transformer rated capacity according to the actual capacity of the transformer; calculating the electrical parameter anomaly coefficient based on the harmonic order, harmonic amplitude, grid frequency fluctuation anomaly coefficient and the transformer load rate based on the preset electrical parameter anomaly formula.

[0008] In a preferred embodiment, the equipment operation abnormality coefficient is specifically obtained by the following method: obtaining the oil temperature of the equipment's transformer, the surface temperature of the equipment, and the vibration signal of the equipment; filtering the vibration signal of the equipment to remove high-frequency noise and low-frequency trends to obtain the vibration frequency of the equipment; performing frequency domain analysis on the vibration frequency of the equipment to obtain the amplitude of the vibration frequency, and calculating the vibration abnormality frequency coefficient of the equipment based on a preset frequency fluctuation stability coefficient using a standard deviation method; obtaining the aging rate and number of operations of the equipment based on regression analysis based on the historical operation data of the substation equipment; calculating the equipment operation abnormality coefficient based on the oil temperature of the equipment's transformer, the surface temperature of the equipment, and the vibration abnormality frequency coefficient in combination with the aging rate and number of operations of the equipment based on a preset equipment operation abnormality formula.

[0009] In a preferred embodiment, the environmental anomaly coefficient is specifically obtained by the following method: monitor and obtain the first data of environmental impact around the substation, the first data including humidity, air pressure, environmental corrosion rate, and frequency and wavelength of electromagnetic waves; obtain the anti-interference and anti-corrosion capabilities of the substation itself; obtain the magnetic field action distance by measuring the distance between the substation and the electromagnetic interference source; obtain the electromagnetic interference intensity based on the radiation power according to the frequency and wavelength of the electromagnetic wave combined with the magnetic field action distance; calculate the environmental anomaly coefficient based on a preset environmental anomaly formula by combining the first data of the substation environment impact and the anti-interference and anti-corrosion capabilities of the substation itself with the electromagnetic interference intensity.

[0010] In a preferred embodiment, the electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient are used to construct a substation partial discharge abnormality model based on machine learning, and the substation partial discharge abnormality coefficient is generated according to the substation partial discharge abnormality model. The specific steps are as follows: collecting partial discharge data and historical electrical parameter abnormality coefficients, historical equipment operation abnormality coefficients and historical environmental abnormality coefficients, the partial discharge data including the intensity of partial discharge, the frequency of partial discharge and the location data of partial discharge, and performing data cleaning and correction data outlier processing; dividing the processed data into a training set for training the model based on machine learning, adjusting the model parameters, assigning weights to the electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient respectively, and constructing a substation partial discharge abnormality model; inputting the newly acquired electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient into the substation partial discharge abnormality model to generate the substation partial discharge abnormality coefficient.

[0011] In a preferred embodiment, the substation partial discharge abnormality coefficient is compared and analyzed based on the preset historical substation partial discharge stability coefficient to determine whether the substation partial discharge situation is faulty and perform early warning processing. The specific steps are as follows: collect the historical substation partial discharge stability coefficient of the substation under extreme working conditions; obtain the substation partial discharge abnormality coefficients at different time nodes, and calculate the average value according to the discrete degree of the substation partial discharge abnormality coefficients at different time nodes. ; The average value Compare and analyze the partial discharge stability coefficient of historical substations; if the average value is less than the historical substation partial discharge stability coefficient, then the substation is working normally at this time; if the average value If it is greater than the historical substation partial discharge stability coefficient, then the substation at this time has a partial discharge abnormal fault and an alarm is issued.

[0012] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By building a substation partial discharge anomaly model through machine learning, it is possible to efficiently identify potential faults in a data-driven manner, improve fault detection accuracy and response speed, reduce labor costs and improve equipment operation reliability.

[0013] 2. Generate fault warnings through artificial intelligence, use automation to analyze massive data, improve prediction accuracy, and take measures in advance, thereby effectively reducing downtime and operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A structural diagram of a substation partial discharge fault early warning method based on artificial intelligence provided in an embodiment of the present application.

[0015] Figure 2A schematic diagram of the structure of an artificial intelligence substation partial discharge fault warning system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0017] Embodiment 1, Figure 1 The schematic diagram of the structure of the artificial intelligence substation partial discharge fault early warning method provided in the embodiment of the present application includes the following steps: S1, obtaining abnormal partial discharge data of a substation, and performing feature extraction on the abnormal partial discharge data of the substation to obtain abnormal coefficients of electrical parameters, abnormal coefficients of equipment operation, and abnormal coefficients of environment.

[0018] The electrical parameter abnormality coefficient refers to the situation in which certain key electrical parameters exceed their normal range or fluctuate abnormally during the operation of the power system or electrical equipment in the substation, which may indicate equipment failure, performance degradation, or potential safety risks. These abnormal data usually reflect the abnormal state of electrical equipment or power network, and can provide important clues for fault diagnosis, early warning and maintenance decision-making.

[0019] Obtaining electrical parameter anomaly coefficients has the following advantages for early warning of partial discharge faults in substations: Early fault detection: Monitoring of abnormal coefficients can identify potential faults in substation equipment in advance, especially the occurrence of partial discharge (PD). PD is usually a signal of deterioration of insulation materials in electrical equipment (such as transformers, cables, etc.), and when PD occurs, electrical parameters (such as current, voltage, frequency, etc.) will fluctuate abnormally. By monitoring these abnormal changes, early warning can be issued before equipment faults further deteriorate, and maintenance or repair measures can be taken to avoid serious damage or shutdown of equipment.

[0020] Improve equipment reliability: By monitoring the abnormal coefficient of electrical parameters, dynamic evaluation of the operating status of substation equipment can be achieved. When there are signs of faults such as partial discharge, corresponding maintenance measures can be taken immediately to avoid the expansion of faults, extend the service life of equipment, and improve the overall reliability of the substation.

[0021] Accurately locate the source of the fault: Abnormal coefficient changes in electrical parameters can provide early information about the occurrence of faults and help locate the problem area. By comparing electrical parameters under different equipment and different operating environments, the type and location of the fault can be determined, thereby performing more accurate fault diagnosis and troubleshooting, and reducing downtime caused by equipment failure.

[0022] Save maintenance costs: Regular monitoring of electrical parameter abnormality coefficients can help substation operators promptly discover potential equipment failure risks. This "preventive maintenance" can avoid high maintenance costs caused by sudden equipment failures, and can also reduce the impact of sudden equipment downtime on production and supply chains. Compared with traditional regular inspections, relying on data monitoring can achieve more accurate maintenance plans and resource allocation, saving a lot of manual inspection costs and emergency maintenance costs.

[0023] Real-time monitoring and automatic alarm: The acquisition and monitoring of electrical parameter abnormality coefficients can achieve real-time monitoring. During the operation of the equipment, the system can automatically alarm according to the set threshold to prompt the staff of abnormal conditions, thereby avoiding problems missed by manual inspections. This automated monitoring system can significantly improve the operation and maintenance efficiency and response speed of the substation.

[0024] Data-driven fault trend analysis: The long-term accumulated electrical parameter abnormal data can be used for trend analysis and fault prediction. By analyzing historical data, the potential failure modes or patterns of certain equipment or systems can be identified, and early warning can be given to avoid production interruptions caused by sudden equipment failures. Compared with traditional manual inspections, data-driven early warning systems have higher accuracy and foresight.

[0025] Optimize operation and maintenance decisions: The electrical parameter abnormality coefficient can provide real-time feedback on the operating status of substation equipment. These data can help decision makers make more scientific and accurate operation and maintenance decisions, arrange maintenance and replacement work reasonably, avoid ineffective or excessive maintenance, and thus achieve optimal allocation of resources.

[0026] Enhanced safety: For substations, electrical faults such as partial discharge may not only damage equipment, but also cause serious accidents such as fire and explosion. By identifying electrical parameter anomalies early, substations can avoid these potential safety hazards, improve overall safety, and ensure the safety of personnel and equipment.

[0027] The specific method for obtaining the electrical parameter abnormality coefficient is as follows: Obtain voltage and current waveforms, actual transformer capacity, and transient waveform data of substation equipment; The voltage and current waveforms of the substation equipment are obtained based on the fast Fourier transform algorithm to obtain the harmonic order and harmonic amplitude; The substation working time period G is divided into N number of nodes, and the grid frequency fluctuation data of each node in the G time period is monitored and obtained; Obtaining a grid frequency fluctuation deviation value of each node according to the grid frequency fluctuation data based on preset normal grid frequency fluctuation data; The power grid frequency fluctuation anomaly coefficient is obtained by performing data processing on the power grid frequency fluctuation deviation value of each node; The load rate of the transformer is calculated based on the preset rated capacity of the transformer according to the actual capacity of the transformer; The electrical parameter abnormality coefficient is calculated based on the preset electrical parameter abnormality formula according to the harmonic order, harmonic amplitude, grid frequency fluctuation abnormality coefficient and transformer load rate.

[0028] The specific calculation formula of the abnormal coefficient of power grid frequency fluctuation is as follows:

[0029] N is N nodes, is the grid frequency fluctuation data of the i-th node, where i=1,2,3,...,R, R is an integer, It is the preset normal grid frequency fluctuation data.

[0030] The specific calculation method of the electrical parameter abnormality coefficient is as follows:

[0031] In the formula, is the electrical parameter anomaly coefficient, is the power grid frequency fluctuation abnormal coefficient, is the load factor of the transformer, is the transient waveform data, is the harmonic order, is the harmonic amplitude N for N nodes, is the grid frequency fluctuation data of the i-th node, where i=1,2,3,...,R, R is an integer, It is the preset normal grid frequency fluctuation data.

[0032] It should be noted that obtaining the abnormal coefficient of grid frequency fluctuation can better predict the electrical failure or partial discharge phenomenon of equipment. Changes in load rate will aggravate the partial discharge problem, especially in the case of overload or load imbalance.

[0033] Equipment operation abnormality coefficient is an indicator used to evaluate whether there are abnormalities or potential faults in substation equipment (such as transformers, cables, switchgear, etc.). It usually evaluates faults in the substation partial discharge process by monitoring the abnormal partial discharge data in the equipment.

[0034] Obtaining equipment operation abnormality coefficients for early warning of substation partial discharge faults has the following advantages: Identify potential faults in advance and reduce equipment damage: The equipment operation abnormality coefficient can reflect the health and operating status of the equipment, especially the hidden fault phenomenon of partial discharge (PD). By monitoring the abnormality coefficient in real time, it is possible to identify possible equipment faults or hidden dangers in advance. For example, partial discharge faults usually do not directly cause equipment failure in the early stages, but if they are not discovered and handled in time, they may aggravate the damage. Therefore, early detection of equipment abnormalities can avoid the deterioration of faults and reduce the degree of equipment damage and maintenance costs.

[0035] Improve operation and maintenance efficiency and optimize maintenance resource allocation: Using abnormal coefficient monitoring can help operation and maintenance personnel locate the source of the fault more accurately and optimize the maintenance plan. In the past, traditional detection methods often relied on regular inspections and manual judgments. By monitoring the abnormal coefficient in real time, more accurate early warnings can be achieved, avoiding unnecessary over-maintenance or ignoring minor problems, thereby rationally allocating maintenance resources and improving operation and maintenance efficiency.

[0036] Reduce the risk of power outages and ensure power supply safety: If a partial discharge fault occurs in the substation equipment, it may cause the equipment to suddenly stop operating or a serious fault occurs, which in turn affects the stability of the entire power system. By monitoring the abnormal coefficient, possible partial discharge faults can be predicted in advance and handled in a timely manner, thereby reducing the risk of sudden power outages, ensuring the stable operation of the power system, and improving power supply reliability.

[0037] Improve the accuracy of fault warning: The abnormal coefficient is obtained by collecting and analyzing data in real time through a variety of sensors (such as voltage, current, temperature, partial discharge signals, etc.). These data can help operation and maintenance personnel more accurately determine whether the equipment is in an abnormal operating state. Compared with the traditional regular inspection method, the early warning system based on data analysis can improve the accuracy of fault prediction and reduce false alarms and missed alarms.

[0038] Enhanced intelligent monitoring capabilities: With the introduction of artificial intelligence and big data technologies, intelligent management of substation equipment can be achieved through data-driven abnormal coefficient monitoring. The system can automatically analyze the historical data of the equipment, identify trend changes, and perform predictive maintenance, helping operation and maintenance personnel to achieve the transition from "passive maintenance to active maintenance" and improve management efficiency.

[0039] Improve the safety of fault handling: Partial discharge faults may lead to major safety hazards such as fire and electrical accidents. Through real-time monitoring and changes in abnormal coefficients, operation and maintenance personnel can take timely measures to prevent partial discharge from developing into serious safety problems, improve the safety of equipment operation and maintenance, and ensure the safety of personnel and equipment.

[0040] Support equipment lifecycle management: In the long-term operation of equipment, monitoring the equipment operation abnormality coefficient not only helps to provide immediate fault warning, but also accumulates historical data on equipment health management. These data can support equipment lifecycle management, help with asset evaluation, optimize equipment replacement plans, and provide a basis for subsequent investment decisions.

[0041] The specific method for obtaining the equipment operation abnormality coefficient is as follows: Obtain the oil temperature of the equipment's transformer, the surface temperature of the equipment, and the vibration signal of the equipment; Filtering the vibration signal of the device to remove high-frequency noise and low-frequency trends to obtain the vibration frequency of the device; The vibration frequency of the equipment is analyzed in the frequency domain to obtain the amplitude of the vibration frequency, and the vibration abnormal frequency coefficient of the equipment is calculated based on the preset frequency fluctuation stability coefficient by the standard deviation method; Based on the historical operation data of substation equipment, the aging rate and operation times of the equipment are obtained through regression analysis; The equipment operation abnormality coefficient is calculated based on the preset equipment operation abnormality formula according to the oil temperature of the equipment's transformer, the surface temperature of the equipment and the vibration abnormality frequency coefficient in combination with the equipment's aging rate and the number of operations.

[0042] The specific calculation formula of the equipment operation abnormality coefficient is as follows:

[0043] In the formula, To prepare for abnormal operation coefficient, is the vibration abnormal frequency coefficient of the equipment, is the aging rate of the equipment, c is the number of times the equipment is run, is the oil temperature of the transformer, is the surface temperature of the equipment, and T is the operating time of the equipment.

[0044] It should be noted that the abnormal vibration frequency coefficient of the equipment can reveal internal faults or damage. When the abnormal vibration frequency coefficient of the equipment becomes larger and larger, the abnormal partial discharge of the substation will become more and more obvious. The aging rate and number of operations of the equipment can help determine whether the equipment is in a high-risk range of failure, which plays an important role in obtaining abnormal equipment operation data.

[0045] Environmental anomaly coefficients refer to those external factors that affect the partial discharge phenomenon. Changes in these environmental factors can have an adverse effect on the operation of the equipment, leading to equipment failure or safety hazards. These include some external climate changes, especially storms and lightning weather, which may cause partial discharge to occur or intensify, and high concentrations of dust, smoke or other pollutants may form a conductive layer on the surface of the equipment, reducing the insulation performance of the equipment, and then causing partial discharge. In addition, electromagnetic interference around the substation (such as radiation from adjacent high-voltage lines or large electrical equipment) may affect the detection accuracy of partial discharge, or indirectly affect the operating status of substation equipment, causing substation partial discharge failure.

[0046] Obtaining the environmental anomaly coefficient has the following advantages for early warning of partial discharge faults in substations: Improve the accuracy of fault warning: The impact of environmental factors on partial discharge. The occurrence of partial discharge is greatly affected by environmental factors (such as temperature, humidity, air pressure, external electromagnetic interference, etc.). By real-time monitoring of environmental parameters and calculating environmental anomaly coefficients, it can help to more accurately identify and judge the abnormal mode of partial discharge. For example, in a high humidity environment, equipment is more prone to partial discharge. Combined with environmental data analysis, it can effectively improve the sensitivity and accuracy of partial discharge detection. Reduce the false alarm rate. Since environmental changes will directly affect the manifestation of partial discharge, judgment through comprehensive environmental data can avoid misjudgment of partial discharge signals caused by environmental fluctuations and reduce false alarms or missed alarms caused by environmental changes.

[0047] Provide earlier fault warning: Discover potential fault hazards in advance. The environmental anomaly coefficient can reflect the operating status of the equipment under specific environmental conditions in real time. If the environmental parameters are abnormal (such as a sudden rise in temperature or a sharp increase in humidity, etc.), it may mean that the stress of equipment operation has increased and the risk of partial discharge has increased. Timely identification of these changes can help predict faults in advance and take measures to prevent equipment damage or shutdown. Dynamically adjust the warning threshold. Depending on different environmental conditions, the detection standard and warning threshold of partial discharge may need to be adjusted dynamically. For example, in a high temperature environment, the insulation performance of the equipment may have partially degraded, and partial discharge may be more likely to occur. In this case, the environmental anomaly coefficient can be used to adjust the fault tolerance range of the equipment and issue an early alarm.

[0048] Reduce human intervention and maintenance costs: Intelligent early warning, by comprehensively considering the impact of environmental factors, the automation system can achieve intelligent monitoring and fault diagnosis. When the environmental parameters reach abnormal values, the system will automatically issue a fault warning, thus avoiding delays or negligence in human intervention. In this way, maintenance personnel can identify potential partial discharge problems in the shortest possible time and take appropriate measures. Optimize operation and maintenance decisions, the environmental anomaly coefficient provides a comprehensive and scientific basis for judgment for operation and maintenance personnel, reduces excessive maintenance, and avoids unnecessary downtime for maintenance. Through accurate early warning, operation and maintenance personnel can intervene in advance when equipment problems occur, reducing the maintenance cost and downtime of equipment failures.

[0049] Improve equipment reliability and operational stability: Dynamically adapt to environmental changes. The equipment operating environment of the substation will change with the seasons. With the help of the environmental anomaly coefficient, the equipment management system can dynamically adjust the equipment's operating conditions and protection strategies according to real-time environmental changes. This helps to maintain high equipment reliability when environmental conditions change. Reduce the occurrence of sudden failures. Abnormal changes in environmental factors may cause performance degradation or aging of certain parts of the equipment. By monitoring these environmental changes and their impact on partial discharge, potential failures can be identified in a timely manner, reducing the occurrence of sudden failures caused by environmental factors and improving the stability of substation equipment.

[0050] Achieve predictive maintenance of equipment health status: More accurate life prediction, the environmental anomaly coefficient can reveal the operating status of the equipment in different environments, so as to make more accurate predictions on the health status and service life of the equipment. For example, equipment that has been in a high temperature and high humidity environment for a long time may be more prone to partial discharge than equipment in a normal environment. Environmental data can be used to more scientifically evaluate the health status of the equipment. Optimize maintenance timing. Through environmental sensitivity analysis of partial discharge of equipment, a smarter maintenance plan can be implemented to avoid excessive maintenance or delayed maintenance, and achieve "maintenance on demand". This not only helps to extend the service life of the equipment, but also avoids equipment failures caused by ignoring environmental factors.

[0051] Enhance the spatial and temporal resolution of partial discharge monitoring: Multi-dimensional data fusion, by combining environmental data and partial discharge signal analysis, the monitoring system can obtain more dimensional information, thereby evaluating the occurrence of partial discharge from multiple angles. For example, combining temperature and humidity data with the changing trend of partial discharge signals can help accurately determine whether partial discharge is caused by environmental factors or equipment problems. To adapt to changes under different environmental conditions, the partial discharge monitoring system of the substation can be personalized for different environmental conditions (such as high humidity, large temperature difference, etc.), improving the system's response sensitivity and early warning capabilities in various environments.

[0052] Improve safety: Environmental factors catalyze partial discharge. Some harsh environmental conditions (such as excessive humidity or high temperature) may accelerate the partial discharge process and even cause equipment failure. By monitoring these abnormal environmental changes and triggering fault warnings and treatment measures in a timely manner, safety issues caused by environmental factors can be effectively avoided, such as electrical fires or short circuit accidents caused by equipment failure. By optimizing the safety assessment model, environmental abnormality data can be used as an important reference for evaluating the safety of substation operation, helping to achieve a more comprehensive and real-time risk assessment, thereby ensuring the safety of operators and equipment.

[0053] The specific method for obtaining the environmental anomaly coefficient is as follows: Monitor and obtain first data of environmental impact around the substation, the first data including humidity, air pressure, environmental corrosion rate, and frequency and wavelength of electromagnetic waves; Obtain the substation's own anti-interference and anti-corrosion capabilities; The magnetic field action distance is obtained by measuring the distance between the substation and the electromagnetic interference source; The electromagnetic interference intensity is obtained based on the frequency and wavelength of the electromagnetic wave combined with the magnetic field action distance and the radiation power; The first data of the substation environmental impact and the anti-interference ability and anti-corrosion ability of the substation itself are combined with the electromagnetic interference intensity to calculate the environmental anomaly coefficient based on a preset environmental anomaly formula.

[0054] The specific calculation method of the environmental anomaly coefficient is as follows:

[0055] In the formula, is the environmental anomaly coefficient, is the electromagnetic interference intensity, For the anti-interference ability of the substation itself, is the environmental corrosion rate, For corrosion resistance, For humidity, For air pressure.

[0056] S2, constructing a substation partial discharge abnormality model based on machine learning using the electrical parameter abnormality coefficient, equipment operation abnormality coefficient and environmental abnormality coefficient, and generating a substation partial discharge abnormality coefficient according to the substation partial discharge abnormality model, the specific steps are as follows: Collecting partial discharge data and historical electrical parameter abnormality coefficients, historical equipment operation abnormality coefficients and historical environmental abnormality coefficients, wherein the partial discharge data includes the intensity, frequency and location data of partial discharge, and performing data cleaning and correction of data abnormality processing; The processed data is divided into training sets for training models based on machine learning, the model parameters are adjusted, the electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient are given weights respectively, and the substation partial discharge abnormality model is constructed; The newly acquired electrical parameter abnormality coefficient, equipment operation abnormality coefficient and environmental abnormality coefficient are input into the substation partial discharge abnormality model to generate the substation partial discharge abnormality coefficient.

[0057] The specific calculation formula of the substation partial discharge abnormal model is as follows:

[0058] In the formula, is the substation partial discharge anomaly coefficient, is the electrical parameter anomaly coefficient, is the environmental anomaly coefficient, is the equipment operation abnormality coefficient, , , is the weight coefficient after training.

[0059] It should be noted that when the electrical parameter abnormality coefficient becomes larger and larger, the environmental abnormality coefficient becomes larger and larger, and the equipment operation abnormality coefficient becomes larger and larger, the substation partial discharge abnormality coefficient will become more and more obvious.

[0060] S3, compare and analyze the substation partial discharge abnormal coefficient based on the preset historical substation partial discharge stability coefficient, determine whether the substation partial discharge situation has a fault and perform early warning processing. The specific steps are as follows: Collect the historical substation partial discharge stability coefficients under extreme working conditions; The substation partial discharge abnormality coefficients at different time nodes are obtained, and the average value is calculated according to the degree of dispersion of the substation partial discharge abnormality coefficients at different time nodes. ; The average Compare and analyze the partial discharge stability coefficients of historical substations; If the average If it is smaller than the partial discharge stability coefficient of the historical substation, the substation is working normally at this time; If the average If it is greater than the historical substation partial discharge stability coefficient, the substation will have a partial discharge abnormal fault, an alarm will be issued, and relevant maintenance personnel will be reminded to troubleshoot and repair the fault to ensure the normal operation of the substation.

[0061] Embodiment 2, Figure 2The schematic diagram of the structure of the artificial intelligence substation partial discharge fault warning system provided in the embodiment of the present application includes a partial discharge abnormal data acquisition module, a partial discharge abnormal model construction module, and a partial discharge abnormal processing module, and there are connections between the modules: A partial discharge abnormal data acquisition module is used to acquire the partial discharge abnormal data of the substation, and perform feature extraction on the partial discharge abnormal data of the substation to obtain an electrical parameter abnormal coefficient, an equipment operation abnormal coefficient, and an environmental abnormal coefficient; A partial discharge abnormal model construction module is used to construct a substation partial discharge abnormal model based on machine learning using the electrical parameter abnormal coefficient, equipment operation abnormal coefficient and environmental abnormal coefficient, and generate a substation partial discharge abnormal coefficient according to the substation partial discharge abnormal model; The partial discharge abnormality processing module is used to compare and analyze the substation partial discharge abnormality coefficient based on the preset historical substation partial discharge stability coefficient, determine whether the substation partial discharge situation has a fault and perform early warning processing.

[0062] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0063] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0064] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0065] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0066] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0067] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A substation partial discharge fault early warning method based on artificial intelligence, characterized in that: The steps include: Acquire abnormal partial discharge data of the substation, and perform feature extraction on the abnormal partial discharge data of the substation to obtain abnormal coefficients of electrical parameters, abnormal coefficients of equipment operation, and abnormal coefficients of environment; The electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient are used to construct a substation partial discharge abnormality model based on machine learning, and the substation partial discharge abnormality coefficient is generated according to the substation partial discharge abnormality model; The substation partial discharge abnormal coefficient is compared and analyzed based on the preset historical substation partial discharge stability coefficient to determine whether the substation partial discharge situation has a fault and take early warning measures.

2. The artificial intelligence-based substation partial discharge fault early warning method according to claim 1 is characterized in that: The specific method for obtaining the electrical parameter abnormality coefficient is as follows: Obtain voltage and current waveforms, actual transformer capacity, and transient waveform data of substation equipment; The voltage and current waveforms of the substation equipment are obtained based on the fast Fourier transform algorithm to obtain the harmonic order and harmonic amplitude; The substation working time period G is divided into N number of nodes, and the grid frequency fluctuation data of each node in the G time period is monitored and obtained; Obtaining a grid frequency fluctuation deviation value of each node based on the grid frequency fluctuation data and preset normal grid frequency fluctuation data; The power grid frequency fluctuation anomaly coefficient is obtained by performing data processing on the power grid frequency fluctuation deviation value of each node; The load rate of the transformer is calculated based on the preset rated capacity of the transformer according to the actual capacity of the transformer; The electrical parameter abnormality coefficient is calculated based on the preset electrical parameter abnormality formula according to the harmonic order, harmonic amplitude, grid frequency fluctuation abnormality coefficient and transformer load rate.

3. The substation partial discharge fault early warning method based on artificial intelligence according to claim 1 is characterized in that: The specific method for obtaining the equipment operation abnormality coefficient is as follows: Obtain the oil temperature of the equipment's transformer, the surface temperature of the equipment, and the vibration signal of the equipment; Filtering the vibration signal of the device to remove high-frequency noise and low-frequency trends to obtain the vibration frequency of the device; The vibration frequency of the equipment is analyzed in the frequency domain to obtain the amplitude of the vibration frequency, and the vibration abnormal frequency coefficient of the equipment is calculated based on the preset frequency fluctuation stability coefficient by the standard deviation method; Based on the historical operation data of substation equipment, the aging rate and operation times of the equipment are obtained through regression analysis; The equipment operation abnormality coefficient is calculated based on the preset equipment operation abnormality formula according to the oil temperature of the equipment's transformer, the surface temperature of the equipment and the vibration abnormality frequency coefficient in combination with the equipment's aging rate and the number of operations.

4. The substation partial discharge fault early warning method based on artificial intelligence according to claim 1 is characterized in that: The specific method for obtaining the environmental anomaly coefficient is as follows: Monitor and obtain first data of environmental impact around the substation, the first data including humidity, air pressure, environmental corrosion rate, and frequency and wavelength of electromagnetic waves; Obtain the substation's own anti-interference and anti-corrosion capabilities; The magnetic field action distance is obtained by measuring the distance between the substation and the electromagnetic interference source; The electromagnetic interference intensity is obtained based on the frequency and wavelength of the electromagnetic wave combined with the magnetic field action distance and the radiation power; The first data of the substation environmental impact and the anti-interference ability and anti-corrosion ability of the substation itself are combined with the electromagnetic interference intensity to calculate the environmental anomaly coefficient based on a preset environmental anomaly formula.

5. The substation partial discharge fault early warning method based on artificial intelligence according to claim 1 is characterized in that: The electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient are used to construct a substation partial discharge abnormality model based on machine learning, and the substation partial discharge abnormality coefficient is generated according to the substation partial discharge abnormality model. The specific steps are as follows: Collecting partial discharge data and historical electrical parameter abnormality coefficients, historical equipment operation abnormality coefficients and historical environmental abnormality coefficients, wherein the partial discharge data includes the intensity, frequency and location data of partial discharge, and performing data cleaning and correction of data abnormality processing; The processed data is divided into training sets for training models based on machine learning, the model parameters are adjusted, the electrical parameter abnormality coefficient, the equipment operation abnormality coefficient and the environmental abnormality coefficient are given weights respectively, and the substation partial discharge abnormality model is constructed; The newly acquired electrical parameter abnormality coefficient, equipment operation abnormality coefficient and environmental abnormality coefficient are input into the substation partial discharge abnormality model to generate the substation partial discharge abnormality coefficient.

6. The substation partial discharge fault early warning method based on artificial intelligence according to claim 1 is characterized in that: The substation partial discharge abnormal coefficient is compared and analyzed based on the preset historical substation partial discharge stability coefficient to determine whether the substation partial discharge situation is faulty and to perform early warning processing. The specific steps are as follows: Collect the historical substation partial discharge stability coefficients under extreme working conditions; The substation partial discharge abnormal coefficients at different time nodes are obtained, and the average value is calculated according to the degree of dispersion of the substation partial discharge abnormal coefficients at different time nodes. ; The average Compare and analyze the partial discharge stability coefficients of historical substations; If the average If it is smaller than the partial discharge stability coefficient of the historical substation, the substation is working normally at this time; If the average If it is greater than the historical substation partial discharge stability coefficient, then the substation at this time has a partial discharge abnormal fault and an alarm is issued.

7. The substation partial discharge fault early warning method based on artificial intelligence according to claim 2 is characterized in that: The specific calculation formula of the abnormal coefficient of power grid frequency fluctuation is as follows: N is N nodes, is the grid frequency fluctuation data of the i-th node, where i=1,2,3,...,R, R is an integer, It is the preset normal power grid frequency fluctuation data; The specific calculation method of the electrical parameter abnormality coefficient is as follows: In the formula, is the electrical parameter anomaly coefficient, is the power grid frequency fluctuation abnormal coefficient, is the load factor of the transformer, is the transient waveform data, is the harmonic order, is the harmonic amplitude N for N nodes, is the grid frequency fluctuation data of the i-th node, where i=1,2,3,...,R, R is an integer, It is the preset normal grid frequency fluctuation data.

8. The substation partial discharge fault early warning method based on artificial intelligence according to claim 3 is characterized in that: The specific calculation formula of the equipment operation abnormality coefficient is as follows: In the formula, To prepare for abnormal operation coefficient, is the vibration abnormal frequency coefficient of the equipment, is the aging rate of the equipment, c is the number of times the equipment is run, is the oil temperature of the transformer, is the surface temperature of the equipment, and T is the operating time of the equipment.

9. The artificial intelligence-based substation partial discharge fault early warning method according to claim 4 is characterized in that: The specific calculation method of the environmental anomaly coefficient is as follows: In the formula, is the environmental anomaly coefficient, is the electromagnetic interference intensity, For the anti-interference ability of the substation itself, is the environmental corrosion rate, For corrosion resistance, For humidity, For air pressure.

10. A system using the artificial intelligence-based substation partial discharge fault early warning method according to any one of claims 1 to 9, characterized in that: It includes a partial discharge abnormal data acquisition module, a partial discharge abnormal model construction module, and a partial discharge abnormal processing module. There are connections between the modules: A partial discharge abnormal data acquisition module is used to acquire the partial discharge abnormal data of the substation, and perform feature extraction on the partial discharge abnormal data of the substation to obtain an electrical parameter abnormal coefficient, an equipment operation abnormal coefficient, and an environmental abnormal coefficient; A partial discharge abnormal model construction module is used to construct a substation partial discharge abnormal model based on machine learning using the electrical parameter abnormal coefficient, equipment operation abnormal coefficient and environmental abnormal coefficient, and generate a substation partial discharge abnormal coefficient according to the substation partial discharge abnormal model; The partial discharge abnormality processing module is used to compare and analyze the substation partial discharge abnormality coefficient based on the preset historical substation partial discharge stability coefficient, determine whether the substation partial discharge situation has a fault and perform early warning processing.

Images