Power distribution station house insulation fault precise diagnosis system and method based on multi-source signal correlation analysis and spatial positioning

Abstract

This invention discloses a precise diagnostic system and method for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning. Specifically, it relates to the field of fault diagnosis technology. The method includes capturing partial discharge signals from electrical equipment using monitoring devices, determining the location of insulation faults in the substation's electrical equipment using an ultrasonic sensor array, and adjusting the environmental conditions of the electrical equipment by modifying the fan operating parameters. It also involves constructing an LSTM predictive partial discharge model for the insulation fault location to generate a predicted partial discharge sequence. Furthermore, it involves collecting UHF signals, ultrasonic signals, and transient ground voltage signals, combined with ambient temperature and humidity, to monitor the partial discharge at the spatial location of the insulation fault in real time, determine the objective function of the fan operating parameter optimization algorithm, and diagnose the insulation fault location. This invention helps to optimize the partial discharge phenomenon at the insulation fault location through fan optimization, thereby achieving precise diagnosis of the insulation fault location.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology, and more specifically, to a precise diagnostic system and method for insulation faults in power distribution substations based on multi-source signal correlation analysis and spatial positioning. Background Technology

[0002] As the operating time of power equipment increases, especially high-voltage electrical equipment in substations such as transformers and switchgear, the aging of insulation materials is gradually becoming a significant potential cause of equipment failure. Partial discharge (PD) is an early manifestation of insulation failure in electrical equipment, typically occurring when the insulation material is defective, aged, or damp. Partial discharge leads to the gradual deterioration of the insulation material, potentially causing equipment failure. Therefore, monitoring and diagnosing partial discharge is crucial for equipment health management.

[0003] Traditional partial discharge monitoring systems typically rely on methods such as ultra-high frequency signals and ultrasonic signals to detect and locate discharge sources by acquiring signals. These technologies are mainly used to monitor partial discharge activity in equipment, but most fail to consider the impact of environmental factors (such as temperature, humidity, and air circulation) on partial discharge. Existing technologies lack effective environmental control methods to optimize the diagnosis of partial discharge phenomena and cannot achieve accurate diagnosis of insulation fault locations in substations by adjusting fan operating parameters. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a precise diagnostic system and method for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A precise diagnostic system for insulation faults in power distribution rooms based on multi-source signal correlation analysis and spatial positioning is characterized by including a partial discharge monitoring module, an LSTM prediction module, an information analysis module, and an optimization algorithm module, with signal connections between the modules;

[0007] The partial discharge monitoring module is used to monitor and capture partial discharge signals of electrical equipment, determine the location of insulation faults in electrical equipment in the substation through an ultrasonic sensor array, and change the environmental conditions of electrical equipment based on adjusting the fan operating parameters.

[0008] The LSTM prediction module is used to construct an LSTM predictive partial discharge model of the insulation fault location using historical UHF signals, ambient temperature and humidity data, and generate a predicted partial discharge sequence of the insulation fault location.

[0009] The information analysis module is used to collect UHF signals, ultrasonic signals and transient ground voltage signals, and combine them with ambient temperature and humidity to monitor the partial discharge at the spatial location of the insulation fault in real time, and determine the discharge intensity information, discharge frequency information and transient ground voltage information at the location of the insulation fault.

[0010] The optimization algorithm module is used to determine the objective function of the wind turbine operating parameter optimization algorithm through comprehensive analysis of discharge intensity information, discharge frequency information and transient ground voltage information, and to diagnose the location of insulation faults.

[0011] In a preferred embodiment, generating a predicted partial discharge sequence for the insulation fault location includes:

[0012] The logic for obtaining the predicted partial discharge sequence from the LSTM partial discharge prediction model is as follows:

[0013] T1. Obtain the historical UHF signal feature sequence of the insulation fault location, wherein the UHF signal feature sequence includes the time point and intensity of the UHF signal at the insulation fault location, obtain the historical ambient temperature sequence and ambient humidity sequence, and time-align the historical UHF signal feature sequence with the historical ambient temperature sequence and ambient humidity sequence.

[0014] T2. Based on the aligned historical data, a partial discharge prediction model is constructed. The deep learning model of LSTM is used to adapt to the changing patterns of the time series. The historical data is divided into training set, validation set and test set. The input of each training sample is the UHF signal features, temperature and humidity of the past W time steps, and the label is the partial discharge intensity of the next H time steps.

[0015] T3. The mean square error is used to measure the difference between the partial discharge intensity output by the model and the actual intensity. The Adam optimizer is used to adjust the weights and biases of the model. The output of the model is the prediction of the partial discharge intensity within the future monitoring period, that is, the prediction of the partial discharge sequence.

[0016] T4. Based on the ambient temperature, ambient humidity, and UHF signal characteristics of the insulation fault location during the current monitoring period, call the LSTM prediction partial discharge model to generate the prediction partial discharge sequence for future monitoring periods.

[0017] In a preferred embodiment, determining the discharge intensity information, discharge frequency information, and transient ground voltage information of the insulation fault location includes:

[0018] Discharge intensity information is represented by a discharge intensity difference coefficient, discharge frequency information is represented by a discharge frequency deviation coefficient, and transient ground voltage information is represented by a transient ground voltage anomaly coefficient. The discharge intensity difference coefficient, This is the discharge frequency deviation coefficient. This is the transient ground voltage anomaly coefficient.

[0019] In a preferred embodiment, the logic for obtaining the discharge intensity difference coefficient is as follows:

[0020] Using the UHF signal characteristics of the current electrical equipment insulation fault location, ambient temperature, and ambient humidity as input, an LSTM predictive partial discharge model is used to obtain the predicted partial discharge sequence of the insulation fault location. Based on the discharge characteristics of the predicted partial discharge sequence, the discharge intensity characteristics of the predicted partial discharge sequence are used as the predicted discharge intensity of the electrical equipment insulation fault location. The predicted discharge intensity of the electrical equipment insulation fault location is denoted as: , i = 1, 2, 3, ..., I, where I is a positive integer, and i is the number of each predicted partial discharge within the future monitoring period. The discharge characteristics include discharge intensity and discharge intensity frequency.

[0021] By adjusting the operating parameters of the fan, the location of the current electrical equipment insulation fault can be determined in the future. The actual partial discharge sequence is used to extract the discharge characteristics of the actual partial discharge sequence, thereby obtaining the actual discharge intensity at the location of the electrical equipment insulation fault. The actual discharge intensity at the location of the electrical equipment insulation fault is then marked as: n = 1, 2, 3, ..., N, where N is a positive integer and n is the number of each actual partial discharge within the future monitoring period;

[0022] The discharge intensity difference coefficient is calculated using the following formula: .

[0023] In a preferred embodiment, the logic for obtaining the discharge frequency deviation coefficient is as follows:

[0024] An LSTM predictive partial discharge model is used to obtain the predicted partial discharge sequence of insulation fault locations within a future monitoring period. By adjusting the wind turbine's operating parameters, the actual partial discharge sequence of insulation fault locations within the future monitoring period is obtained. The number of partial discharges in the predicted partial discharge sequence is marked as follows: The number of partial discharges in the actual partial discharge sequence is marked as follows: ;

[0025] The discharge frequency deviation coefficient is calculated using the following formula: .

[0026] In a preferred embodiment, the logic for obtaining the transient ground voltage anomaly coefficient is as follows:

[0027] By adjusting the operating parameters of the fan, the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring period is obtained, and the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring period is marked as: Where k = 1, 2, 3, ..., K, K is a positive integer, and k is the number of the transient ground voltage collected during the future monitoring period;

[0028] Set a transient ground voltage intensity threshold, and mark the transient ground voltage intensity threshold as: The transient ground voltage intensity within a future monitoring period is compared with a transient ground voltage intensity threshold. The transient ground voltage values ​​exceeding the threshold within the future monitoring period are then identified and recorded.

[0029] The transient ground voltage of the voltage intensity threshold is relabeled as: Where m = 1, 2, 3, ..., M, M is a positive integer, and m is the transient ground voltage number that is greater than the transient ground voltage intensity threshold.

[0030] The transient ground voltage anomaly coefficient is calculated using the following formula: .

[0031] In a preferred embodiment, determining the objective function of the optimization algorithm includes:

[0032] By comprehensively analyzing the discharge intensity, discharge frequency, and transient ground voltage information at the insulation fault location, and normalizing the discharge intensity difference coefficient, discharge frequency deviation coefficient, and transient ground voltage anomaly coefficient, an adjustment evaluation model is constructed, generating adjustment evaluation coefficients. The expression for the adjustment evaluation coefficients is as follows: ;in, To adjust the evaluation coefficients, These are the proportional coefficients for the discharge intensity difference coefficient, the discharge frequency deviation coefficient, and the transient ground voltage anomaly coefficient, respectively. All are greater than 0.

[0033] In a preferred embodiment, the diagnosis of the location of an insulation fault includes:

[0034] The adjustment evaluation coefficient is used as the objective function of the optimization algorithm, and constraints on the operating parameters are added during the optimization process. By monitoring the operating parameters of the wind turbine, the optimization algorithm adjusts the operating parameters of the wind turbine according to the current value of the adjustment evaluation coefficient. By maximizing the objective function of the optimization algorithm, the maximum adjustment evaluation coefficient for the insulation fault location is obtained, including:

[0035] By monitoring the location of insulation faults, adjustment evaluation coefficients for insulation faults are obtained at different monitoring time periods. Based on the operating parameters of the wind turbine at different monitoring time periods, the operating parameters of the wind turbine are continuously adjusted. The optimization algorithm adjusts the operating parameters of the wind turbine step by step according to the calculated adjustment evaluation coefficients and ensures that they are within the set constraints.

[0036] If the adjustment evaluation coefficient threshold is set, and the adjustment evaluation coefficient of the insulation fault location is always greater than the adjustment evaluation coefficient threshold, the operating parameters of the wind turbine are gradually adjusted until the adjustment evaluation coefficient is maximized, and the insulation fault location is diagnosed as a low-risk insulation fault area. If the adjustment evaluation coefficient of the insulation fault location is greater than the adjustment evaluation coefficient threshold, the optimization algorithm stops optimizing the operating parameters of the wind turbine, and the insulation fault location is diagnosed as a high-risk insulation fault area.

[0037] In a preferred embodiment, the method for accurate diagnosis of insulation faults in substations based on multi-source signal correlation analysis and spatial positioning specifically includes the following steps:

[0038] S1: Real-time monitoring of partial discharge in equipment using ultra-high frequency and ultrasonic signal monitoring devices, and precise location of faults using an ultrasonic sensor array;

[0039] S2: Using historical UHF signals and temperature and humidity data, the partial discharge intensity during the future monitoring period is predicted through an LSTM deep learning model, and a predicted partial discharge sequence is generated.

[0040] S3: Analyze the current insulation fault space health status of the equipment based on discharge intensity, discharge frequency and transient ground voltage signal;

[0041] S4: Construct an adjustment assessment model, generate adjustment assessment coefficients, adjust the wind turbine operating parameters through optimization algorithms, and diagnose the location of insulation faults.

[0042] The technical effects and advantages of this invention are as follows:

[0043] This invention monitors the partial discharge of electrical equipment in real time, utilizing data from various sensors such as UHF signals, ultrasonic signals, and transient ground voltage signals, combined with environmental temperature and humidity information, to analyze the intensity and frequency of partial discharge. By using an LSTM deep learning model to predict discharge trends and optimizing environmental conditions in conjunction with the fan control system, this invention can adjust the operating parameters of the fan in the substation, optimize the partial discharge phenomenon at the location of insulation faults, and thus achieve accurate diagnosis of insulation fault locations. Attached Figure Description

[0044] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0045] Figure 1 This is a schematic diagram of the structure of the present invention;

[0046] Figure 2 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Example 1

[0049] Figure 1 This is a schematic diagram of the structure of the precise diagnosis system for insulation faults in power distribution rooms based on multi-source signal correlation analysis and spatial positioning of the present invention. It includes a partial discharge monitoring module, an LSTM prediction module, an information analysis module, and an optimization algorithm module, with signal connections between the modules.

[0050] The partial discharge monitoring module is used to monitor and capture partial discharge signals of electrical equipment, determine the location of insulation faults in electrical equipment in the substation through an ultrasonic sensor array, and change the environmental conditions of electrical equipment based on adjusting the fan operating parameters.

[0051] The LSTM prediction module is used to construct an LSTM predictive partial discharge model of the insulation fault location using historical UHF signals, ambient temperature and humidity data, and generate a predicted partial discharge sequence of the insulation fault location.

[0052] The information analysis module is used to collect UHF signals, ultrasonic signals and transient ground voltage signals, and combine them with ambient temperature and humidity to monitor the partial discharge at the spatial location of the insulation fault in real time, and determine the discharge intensity information, discharge frequency information and transient ground voltage information at the location of the insulation fault.

[0053] The optimization algorithm module is used to determine the objective function of the wind turbine operating parameter optimization algorithm through comprehensive analysis of discharge intensity information, discharge frequency information and transient ground voltage information, and to diagnose the location of insulation faults.

[0054] In the substation, ultra-high frequency (UHF) and ultrasonic signals are acquired. Since these signals are closely related to partial discharge, monitoring them helps diagnose insulation faults in electrical equipment. UHF signals are common electromagnetic signals during partial discharge, typically ranging from 300 MHz to 3 GHz. Partial discharge sources usually generate UHF signals when the insulation material of electrical equipment is damaged or aged. Ultrasonic signals are mechanical waves generated by partial discharge; they can propagate to the outside of the equipment and be received by ultrasonic sensors. Ultrasonic signals help determine the location and intensity of partial discharges, and time-of-flight (TOF) positioning methods can be used to precisely locate the discharge source.

[0055] It should be noted that the acquisition equipment for ultra-high frequency (UHF) signals is a UHF sensor, installed on the casing, connectors, cable terminals, etc., of electrical equipment, close to potential partial discharge sources, especially in areas where partial discharge may occur. After the UHF signal is received by the sensor, it is usually amplified by a signal amplifier before being transmitted to the data acquisition system. By analyzing the amplitude and frequency changes of the signal, the presence and intensity of the partial discharge can be determined. The acquisition equipment for ultrasonic signals is an ultrasonic sensor, used to capture the sound wave signals caused by partial discharge. It is usually installed outside the equipment. For spatial positioning, multiple ultrasonic sensors need to be deployed, usually in an array. The location of the discharge source can be determined by calculating the time difference of multiple sensors.

[0056] Partial discharge refers to the electrical discharge phenomenon in electrical equipment where the insulation material in a local area undergoes partial breakdown or is about to break down. Partial discharge usually occurs when the insulation material is defective, aged, or damp. Partial discharge often recurs in the same area or location. As the equipment operates, especially under long-term voltage, the occurrence of partial discharge becomes periodic and frequent. Initially, partial discharge may be relatively weak and have a low frequency, but as the discharge point is damaged, the discharge intensity will gradually increase, and the frequency may increase.

[0057] Partial discharge often occurs repeatedly at the same location in electrical equipment. Partial discharge at the same location is an important feature in partial discharge diagnosis, which helps to understand the insulation fault of electrical equipment and the health status of the equipment by observing changes in discharge patterns. Therefore, deploying an ultrasonic sensor array can determine the specific spatial location of the insulation fault.

[0058] The intensity and frequency of partial discharge are often closely related to environmental conditions (such as temperature and humidity). Therefore, in critical areas of a power distribution station, fans should be deployed to mitigate the frequency of partial discharge. These fans should be adjustable in speed to provide appropriate ventilation under different environmental conditions. A partial discharge monitoring system should also be designed, using devices such as UHF sensors and ultrasonic sensors to monitor partial discharge in real time. Once a partial discharge occurs and reaches a certain intensity, the location and intensity of the discharge source should be immediately analyzed and determined. The partial discharge monitoring system is linked to the fan control system; when the system detects an increase in the frequency or intensity of the partial discharge, it can automatically increase the fan's operating frequency, enhance ventilation around the equipment, and reduce equipment temperature and humidity, thereby reducing discharge phenomena.

[0059] For example, power distribution equipment, especially transformers and switchgear, generates heat during operation. If the temperature inside the substation is too high, it accelerates the aging of insulation materials, reduces their insulation performance, and thus increases the frequency and intensity of partial discharge. Fans can effectively cool the equipment, increase air circulation, and lower the temperature around the equipment, thereby reducing the impact of excessive temperature on the equipment's insulation performance. Secondly, fans not only regulate temperature and humidity, but also help remove dust, particulate matter, and other pollutants from the air inside the substation, which can adhere to the surfaces of equipment and increase the risk of partial discharge. Fans, through airflow, can help remove airborne particles, preventing them from settling on equipment surfaces and thus reducing partial discharge caused by pollutants.

[0060] Partial discharge is spatially located using ultrasound to determine the location of the insulation fault. A partial discharge monitoring system is then used to monitor the location of the insulation fault, determining the ambient temperature, humidity, and UHF signal characteristics during the monitoring period. An LSTM predictive partial discharge model for the insulation fault location is then constructed, generating a predicted partial discharge sequence. The logic for obtaining the predicted partial discharge sequence from the LSTM predictive partial discharge model is as follows:

[0061] T1. Obtain the historical UHF signal feature sequence of the insulation fault location, wherein the UHF signal feature sequence includes the time point and intensity of the UHF signal at the insulation fault location, obtain the historical ambient temperature sequence and ambient humidity sequence, and time-align the historical UHF signal feature sequence with the historical ambient temperature sequence and ambient humidity sequence.

[0062] T2. Based on the aligned historical data, a partial discharge prediction model is constructed. The deep learning model of LSTM is used to adapt to the changing patterns of the time series. The historical data is divided into training set, validation set and test set. The input of each training sample is the UHF signal features, temperature and humidity of the past W time steps, and the label is the partial discharge intensity of the next H time steps.

[0063] T3. The mean square error is used to measure the difference between the partial discharge intensity output by the model and the actual intensity. The Adam optimizer is used to adjust the weights and biases of the model. The output of the model is the prediction of the partial discharge intensity within the future monitoring period, that is, the prediction of the partial discharge sequence.

[0064] T4. Based on the ambient temperature, ambient humidity, and UHF signal characteristics of the insulation fault location during the current monitoring period, call the LSTM prediction partial discharge model to generate the prediction partial discharge sequence for future monitoring periods.

[0065] Because the fan has a significant impact on partial discharge at the insulation fault location in the early stage of insulation fault, as the service time of electrical equipment increases, especially after experiencing multiple partial discharges, the fan's effect on reducing partial discharge will gradually decrease. Therefore, by adjusting the fan's operating parameters, based on the LSTM prediction partial discharge model, the discharge intensity information, discharge frequency information, and transient ground voltage information at the insulation fault location are collected. The discharge intensity information is represented by the discharge intensity difference coefficient, the discharge frequency information is represented by the discharge frequency deviation coefficient, and the transient ground voltage information is represented by the transient ground voltage anomaly coefficient.

[0066] Among them, the discharge intensity information and discharge frequency information provide real-time intensity and frequency characteristics of partial discharge, which can reflect the current health status of the insulation fault space of the equipment. If the discharge intensity and discharge frequency are large, it indicates that the partial discharge phenomenon is more serious, and the insulation material may be significantly degraded, which increases the risk of electrical equipment failure. Since the fan operating parameters are designed to optimize environmental factors to control the intensity of partial discharge, when a large partial discharge intensity is detected, the fan operating parameters can be optimized by adjustment. This indicates that the occurrence rate and intensity of partial discharge can be effectively reduced in the early stage of insulation faults.

[0067] The logic for obtaining the discharge intensity difference coefficient is as follows: Using the UHF signal characteristics, ambient temperature, and ambient humidity of the current electrical equipment insulation fault location as input, an LSTM partial discharge prediction model is used to obtain the predicted partial discharge sequence of the insulation fault location. Based on the discharge characteristics of the predicted partial discharge sequence, the discharge intensity characteristics of the predicted partial discharge sequence are used as the predicted discharge intensity of the electrical equipment insulation fault location. The predicted discharge intensity of the electrical equipment insulation fault location is then labeled as follows: , i = 1, 2, 3, ..., I, where I is a positive integer, and i is the number of each predicted partial discharge within the future monitoring period. The discharge characteristics include discharge intensity and discharge intensity frequency.

[0068] By adjusting the operating parameters of the fan, the location of the current electrical equipment insulation fault can be determined in the future. The actual partial discharge sequence is used to extract the discharge characteristics of the actual partial discharge sequence, thereby obtaining the actual discharge intensity at the location of the electrical equipment insulation fault. The actual discharge intensity at the location of the electrical equipment insulation fault is then marked as: n = 1, 2, 3, ..., N, where N is a positive integer and n is the number of each actual partial discharge within the future monitoring period;

[0069] It should be noted that the operating parameters of the fan include the fan speed, fan direction, fan opening degree, and the fan's operating time and cycle during the monitoring period. Higher wind speed may help reduce the temperature and humidity in the substation, which helps reduce the occurrence of partial discharge. The fan direction determines the air flow path, which directly affects how air flows through the substation equipment, and thus affects the area of ​​partial discharge. It also determines the fan's intake air volume, which in turn affects the air flow rate in the substation room.

[0070] The discharge intensity difference coefficient is calculated using the following formula: ;in, This is the discharge intensity difference coefficient.

[0071] As can be seen from the formula, the larger the discharge intensity difference coefficient, the more effectively the partial discharge at the location of the electrical equipment insulation fault has been changed after adjusting the fan's operating parameters. The smaller the discharge intensity, the more the current fan's operating parameters can alleviate the damage caused by the partial discharge at the location of the electrical equipment insulation fault.

[0072] The logic for obtaining the discharge frequency deviation coefficient is as follows: An LSTM predictive partial discharge model is used to obtain the predicted partial discharge sequence of insulation fault locations within a future monitoring period. Then, by adjusting the operating parameters of the wind turbine, the actual partial discharge sequence of insulation fault locations within the future monitoring period is obtained. The number of partial discharges in the predicted partial discharge sequence is marked as follows: The number of partial discharges in the actual partial discharge sequence is marked as follows: ;

[0073] The discharge frequency deviation coefficient is calculated using the following formula: ;in, This is the discharge frequency deviation coefficient.

[0074] It should be noted that the number of partial discharges in the predicted partial discharge sequence represents the number of partial discharge events given by the prediction model, while the actual partial discharge sequence represents the actual number of partial discharge events after adjusting the fan. If the number of partial discharges in the measured partial discharge sequence is greater than the actual number of partial discharge events, it indicates that the fan's operating parameters have optimized the partial discharge problem at the insulation fault location. Conversely, if the number of partial discharges in the measured partial discharge sequence is less than or close to the actual number of partial discharge events, it indicates that adjusting the fan's operating parameters has no effect on the partial discharge at the insulation fault location. This further suggests that the fan's operating parameters may be incorrectly set or the insulation fault of the electrical equipment may be more severe, with less influence from the substation environment.

[0075] As can be seen from the formula, the larger the discharge frequency deviation coefficient, the greater the difference between the predicted and actual number of discharges. This may indicate that the fan adjustment can effectively reduce partial discharges, and that the current fan operating parameters can alleviate the damage caused by partial discharges at the location of electrical equipment insulation faults.

[0076] The advantage of the transient ground voltage anomaly coefficient is that it can clearly show whether the equipment has experienced strong partial discharge during the monitoring period and quantify the intensity of these discharges. As part of a multi-source signal, the transient ground voltage signal, together with the UHF signal, reflects the partial discharge performance at the location of insulation faults in electrical equipment. By combining multiple signal sources, the occurrence, development, and location of partial discharges can be assessed more accurately.

[0077] The logic for obtaining the transient ground voltage anomaly coefficient is as follows: by adjusting the operating parameters of the fan, the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring time period is obtained, and the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring time period is marked as follows: Where k = 1, 2, 3, ..., K, K is a positive integer, and k is the number of the transient ground voltage collected during the future monitoring period;

[0078] Set a transient ground voltage intensity threshold, and mark the transient ground voltage intensity threshold as: The transient ground voltage intensity within a future monitoring period is compared with a transient ground voltage intensity threshold. The transient ground voltage values ​​exceeding the threshold within the future monitoring period are then identified and recorded.

[0079] The transient ground voltage of the voltage intensity threshold is relabeled as: Where m = 1, 2, 3, ..., M, M is a positive integer, and m is the transient ground voltage number that is greater than the transient ground voltage intensity threshold.

[0080] It should be noted that, based on the location of insulation faults, high-frequency voltage signals are captured by voltage sensors. Strong transient ground voltage signals usually correspond to intense partial discharge activity, which can significantly affect the insulation material. Small transient ground voltage signals typically indicate lower intensity and shorter duration of partial discharge. These signals may not be sufficient to cause significant damage to the equipment's insulation, and therefore their impact on equipment performance is relatively small. Thus, if adjusting the fan's operating parameters fails to reduce the transient ground voltage signal, it indicates that the insulation fault location of the electrical equipment stems from a fault or aging of the equipment itself, rather than environmental factors.

[0081] The transient ground voltage anomaly coefficient is calculated using the following formula: ;in, This is the transient ground voltage anomaly coefficient.

[0082] As can be seen from the formula, the larger the transient ground voltage anomaly coefficient, the more intense the partial discharge activity the equipment has experienced, and the stronger the voltage signal of these discharge activities. This may mean that the equipment has poor insulation, and it is necessary to repair or replace the insulation material. Furthermore, adjusting the operating parameters of the fan did not significantly improve the impact of partial discharge.

[0083] By comprehensively analyzing the discharge intensity, discharge frequency, and transient ground voltage information at the insulation fault location, and normalizing the discharge intensity difference coefficient, discharge frequency deviation coefficient, and transient ground voltage anomaly coefficient, an adjustment evaluation model is constructed, generating adjustment evaluation coefficients. The expression for the adjustment evaluation coefficients is as follows: ;in, To adjust the evaluation coefficients, These are the proportional coefficients for the discharge intensity difference coefficient, the discharge frequency deviation coefficient, and the transient ground voltage anomaly coefficient, respectively. All are greater than 0.

[0084] As can be seen from the formula, the larger the discharge intensity difference coefficient and discharge frequency deviation coefficient, the smaller the transient ground voltage anomaly coefficient, and the larger the adjustment evaluation coefficient. This indicates that the intensity of partial discharge changes significantly after adjustment, possibly because the optimization of the fan parameters effectively reduces the discharge intensity and improves the insulation performance of the equipment. Conversely, it indicates that the current operating parameters of the fan have little impact on insulation faults.

[0085] The adjustment evaluation coefficient is used as the objective function of the optimization algorithm. Constraints on operating parameters are added during the optimization process. These operating parameters include the fan speed, fan direction, fan opening degree, and the fan's operating time and cycle within the monitoring period. By monitoring the fan's operating parameters, the optimization algorithm adjusts these parameters based on the current value of the adjustment evaluation coefficient. By maximizing the objective function of the optimization algorithm, the maximum adjustment evaluation coefficient for the insulation fault location is obtained, including:

[0086] By monitoring the location of insulation faults, adjustment evaluation coefficients for insulation faults are obtained at different monitoring time periods. Based on the operating parameters of the wind turbine at different monitoring time periods, the operating parameters of the wind turbine are continuously adjusted. The optimization algorithm adjusts the operating parameters of the wind turbine step by step according to the calculated adjustment evaluation coefficients and ensures that they are within the set constraints.

[0087] If the adjustment evaluation coefficient threshold is set, and the adjustment evaluation coefficient of the insulation fault location is always greater than the adjustment evaluation coefficient threshold, the operating parameters of the fan are gradually adjusted until the adjustment evaluation coefficient is at its maximum. This indicates that adjusting the operating parameters of the fan can effectively improve the partial discharge at the insulation fault location and diagnose the insulation fault location as a low-risk insulation fault area. The electrical equipment does not need to be repaired temporarily and can be improved by the fan. If the adjustment evaluation coefficient of the insulation fault location is greater than the adjustment evaluation coefficient threshold, the optimization algorithm stops optimizing the operating parameters of the fan and diagnoses the insulation fault location as a high-risk insulation fault area. The insulation material of the electrical equipment needs to be replaced and repaired immediately by professionals.

[0088] Example 2

[0089] Figure 2 This is a flowchart illustrating the precise diagnosis method for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning according to the present invention. The method specifically includes the following steps:

[0090] S1: Real-time monitoring of partial discharge in equipment using ultra-high frequency and ultrasonic signal monitoring devices, and precise location of faults using an ultrasonic sensor array;

[0091] S2: Using historical UHF signals and temperature and humidity data, the partial discharge intensity during the future monitoring period is predicted through an LSTM deep learning model, and a predicted partial discharge sequence is generated.

[0092] S3: Analyze the current insulation fault space health status of the equipment based on discharge intensity, discharge frequency and transient ground voltage signal;

[0093] S4: Construct an adjustment assessment model, generate adjustment assessment coefficients, adjust the wind turbine operating parameters through optimization algorithms, and diagnose the location of insulation faults.

[0094] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0095] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0096] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0097] Those skilled in the art will recognize that the units and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0098] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0099] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0100] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A precise diagnostic system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning, characterized in that... It includes a partial discharge monitoring module, an LSTM prediction module, an information analysis module, and an optimization algorithm module, with signal connections between the modules; The partial discharge monitoring module is used to monitor and capture partial discharge signals of electrical equipment, determine the location of insulation faults in electrical equipment in the substation through an ultrasonic sensor array, and change the environmental conditions of electrical equipment based on adjusting the fan operating parameters. The LSTM prediction module is used to construct an LSTM predictive partial discharge model of the insulation fault location using historical UHF signals, ambient temperature and humidity data, and generate a predicted partial discharge sequence of the insulation fault location. The information analysis module is used to collect UHF signals, ultrasonic signals and transient ground voltage signals, and combine them with ambient temperature and humidity to monitor the partial discharge at the spatial location of the insulation fault in real time, and determine the discharge intensity information, discharge frequency information and transient ground voltage information at the location of the insulation fault. The optimization algorithm module is used to determine the objective function of the wind turbine operating parameter optimization algorithm through comprehensive analysis of discharge intensity information, discharge frequency information and transient ground voltage information, and to diagnose the location of insulation faults.

2. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 1, characterized in that... Generate a predicted partial discharge sequence for the location of the insulation fault, including: The logic for obtaining the predicted partial discharge sequence from the LSTM partial discharge prediction model is as follows: T1. Obtain the historical UHF signal feature sequence of the insulation fault location, wherein the UHF signal feature sequence includes the time point and intensity of the UHF signal at the insulation fault location, obtain the historical ambient temperature sequence and ambient humidity sequence, and time-align the historical UHF signal feature sequence with the historical ambient temperature sequence and ambient humidity sequence. T2. Based on the aligned historical data, a partial discharge prediction model is constructed. The deep learning model of LSTM is used to adapt to the changing patterns of the time series. The historical data is divided into training set, validation set and test set. The input of each training sample is the UHF signal features, temperature and humidity of the past W time steps, and the label is the partial discharge intensity of the next H time steps. T3. The mean square error is used to measure the difference between the partial discharge intensity output by the model and the actual intensity. The Adam optimizer is used to adjust the weights and biases of the model. The output of the model is the prediction of the partial discharge intensity within the future monitoring period, that is, the prediction of the partial discharge sequence. T4. Based on the ambient temperature, ambient humidity, and UHF signal characteristics of the insulation fault location during the current monitoring period, call the LSTM prediction partial discharge model to generate the prediction partial discharge sequence for future monitoring periods.

3. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 2, characterized in that, The discharge intensity, discharge frequency, and transient ground voltage information are used to determine the location of the insulation fault, including: Discharge intensity information is represented by a discharge intensity difference coefficient, discharge frequency information is represented by a discharge frequency deviation coefficient, and transient ground voltage information is represented by a transient ground voltage anomaly coefficient. The discharge intensity difference coefficient, This is the discharge frequency deviation coefficient. This is the transient ground voltage anomaly coefficient.

4. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 3, characterized in that, The logic for obtaining the discharge intensity difference coefficient is as follows: Using the UHF signal characteristics of the current electrical equipment insulation fault location, ambient temperature, and ambient humidity as input, an LSTM predictive partial discharge model is used to obtain the predicted partial discharge sequence of the insulation fault location. Based on the discharge characteristics of the predicted partial discharge sequence, the discharge intensity characteristics of the predicted partial discharge sequence are used as the predicted discharge intensity of the electrical equipment insulation fault location. The predicted discharge intensity of the electrical equipment insulation fault location is denoted as: , i = 1, 2, 3, ..., I, where I is a positive integer, and i is the number of each predicted partial discharge within the future monitoring period. The discharge characteristics include discharge intensity and discharge intensity frequency. By adjusting the operating parameters of the fan, the location of the current electrical equipment insulation fault can be determined in the future. The actual partial discharge sequence is used to extract the discharge characteristics of the actual partial discharge sequence, thereby obtaining the actual discharge intensity at the location of the electrical equipment insulation fault. The actual discharge intensity at the location of the electrical equipment insulation fault is then marked as: n = 1, 2, 3, ..., N, where N is a positive integer and n is the number of each actual partial discharge within the future monitoring period; The discharge intensity difference coefficient is calculated using the following formula: .

5. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 4, characterized in that, The logic for obtaining the discharge frequency deviation coefficient is as follows: An LSTM predictive partial discharge model is used to obtain the predicted partial discharge sequence of insulation fault locations within a future monitoring period. By adjusting the wind turbine's operating parameters, the actual partial discharge sequence of insulation fault locations within the future monitoring period is obtained. The number of partial discharges in the predicted partial discharge sequence is marked as follows: The number of partial discharges in the actual partial discharge sequence is marked as follows: ; The discharge frequency deviation coefficient is calculated using the following formula: .

6. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 5, characterized in that, The logic for obtaining the transient ground voltage anomaly coefficient is as follows: By adjusting the operating parameters of the fan, the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring period is obtained, and the transient ground voltage intensity of the electrical equipment insulation fault location within the future monitoring period is marked as: Where k = 1, 2, 3, ..., K, K is a positive integer, and k is the number of the transient ground voltage collected during the future monitoring period; Set a transient ground voltage intensity threshold, and mark the transient ground voltage intensity threshold as: The transient ground voltage intensity within a future monitoring period is compared with a transient ground voltage intensity threshold. The transient ground voltage values ​​exceeding the threshold within the future monitoring period are then identified and recorded. The transient ground voltage of the voltage intensity threshold is relabeled as: Where m = 1, 2, 3, ..., M, M is a positive integer, and m is the transient ground voltage number that is greater than the transient ground voltage intensity threshold. The transient ground voltage anomaly coefficient is calculated using the following formula: .

7. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 6, characterized in that, Determine the objective function of the optimization algorithm, including: By comprehensively analyzing the discharge intensity, discharge frequency, and transient ground voltage information at the insulation fault location, and normalizing the discharge intensity difference coefficient, discharge frequency deviation coefficient, and transient ground voltage anomaly coefficient, an adjustment evaluation model is constructed, generating adjustment evaluation coefficients. The expression for the adjustment evaluation coefficients is as follows: ;in, To adjust the evaluation coefficients, These are the proportional coefficients for the discharge intensity difference coefficient, the discharge frequency deviation coefficient, and the transient ground voltage anomaly coefficient, respectively. All are greater than 0.

8. The precise diagnosis system for insulation faults in substations based on multi-source signal correlation analysis and spatial positioning as described in claim 7, characterized in that, Diagnosis of insulation fault location includes: The adjustment evaluation coefficient is used as the objective function of the optimization algorithm, and constraints on the operating parameters are added during the optimization process. By monitoring the operating parameters of the wind turbine, the optimization algorithm adjusts the operating parameters of the wind turbine according to the current value of the adjustment evaluation coefficient. By maximizing the objective function of the optimization algorithm, the maximum adjustment evaluation coefficient for the insulation fault location is obtained, including: By monitoring the location of insulation faults, adjustment evaluation coefficients for insulation faults are obtained at different monitoring time periods. Based on the operating parameters of the wind turbine at different monitoring time periods, the operating parameters of the wind turbine are continuously adjusted. The optimization algorithm adjusts the operating parameters of the wind turbine step by step according to the calculated adjustment evaluation coefficients and ensures that they are within the set constraints. If the adjustment evaluation coefficient threshold is set, and the adjustment evaluation coefficient of the insulation fault location is always greater than the adjustment evaluation coefficient threshold, the operating parameters of the wind turbine are gradually adjusted until the adjustment evaluation coefficient is maximized, and the insulation fault location is diagnosed as a low-risk insulation fault area. If the adjustment evaluation coefficient of the insulation fault location is greater than the adjustment evaluation coefficient threshold, the optimization algorithm stops optimizing the operating parameters of the wind turbine, and the insulation fault location is diagnosed as a high-risk insulation fault area.

9. A method for accurate diagnosis of insulation faults in substations based on multi-source signal correlation analysis and spatial positioning, used to implement the system described in any one of claims 1-8, characterized in that, Specifically, the following steps are included: S1: Real-time monitoring of partial discharge in equipment using ultra-high frequency and ultrasonic signal monitoring devices, and precise location of faults using an ultrasonic sensor array; S2: Using historical UHF signals and temperature and humidity data, the partial discharge intensity during the future monitoring period is predicted through an LSTM deep learning model, and a predicted partial discharge sequence is generated. S3: Analyze the current insulation fault space health status of the equipment based on discharge intensity, discharge frequency and transient ground voltage signal; S4: Construct an adjustment assessment model, generate adjustment assessment coefficients, adjust the wind turbine operating parameters through optimization algorithms, and diagnose the location of insulation faults.

Images