Location detection method for partial discharge of electrical equipment

Abstract

The invention relates to the technical field of electrical equipment state monitoring, and discloses an electrical equipment partial discharge positioning detection method. The method comprises the following steps: acquiring synchronous current pulse sequences and ultrasonic waveforms through a pulse current sensor and an ultrasonic sensor; calculating a signal delay inequality; establishing a sound wave propagation multi-path database based on the three-dimensional structure of the switch cabinet and the sensor coordinates; matching the ultrasonic shape features with a database to determine an actual propagation path and an equivalent distance; combining the delay inequality, the sensor coordinate and the equivalent distance, and iteratively solving the optimal space coordinate of the discharge point by adopting a particle swarm optimization algorithm; and mapping the coordinates to the topological graph to generate a positioning report. According to the method, the sound wave aliasing effect is decoupled through the pre-established multi-path database, the time delay correction and coordinate solving process is optimized, and the positioning precision and efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical equipment status monitoring, and in particular to a method for locating and detecting partial discharge of electrical equipment. Background Art

[0002] In the field of power equipment condition monitoring, partial discharge positioning is a key technology for diagnosing insulation defects. Due to the complex structure of metal-enclosed switchgear, the internal discharge sound waves form multiple reflections between the inner wall of the cavity and the insulating baffles, resulting in severe aliasing of the ultrasonic signal. Traditional methods rely on manual experience to distinguish between direct waves and reflected waves. The path identification process has significant delays, making it difficult to meet the timeliness requirements of fault warning. In addition, the microsecond delay difference between the pulse current and the ultrasonic signal, if not corrected, will be magnified to a meter-level spatial error due to geometric relationships. Repeated iterative coordinate calculation is required, further limiting positioning efficiency.

[0003] In existing technologies, acoustic wave path analysis often uses real-time simulation of reflection trajectories, and the computational load increases exponentially with cavity complexity. Distance errors caused by time delays require multiple iterations of correction using methods such as gradient descent, which is particularly time-consuming in large-scale substation equipment inspection scenarios. Regional power grid service providers have reported that existing solutions take an average of over 20 minutes to locate a location in switchgear with dense reflection paths, and the results are easily influenced by subjective judgment. A positioning method that can automatically decouple acoustic wave multipath effects and optimize the delay calculation architecture is urgently needed to meet the rapid diagnosis needs of high-density power equipment. Summary of the Invention

[0004] In order to overcome the above-mentioned defects in the prior art, an embodiment of the present invention provides a method for locating and detecting partial discharge of electrical equipment to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above-mentioned object, the present invention provides a method for locating and detecting partial discharge of electrical equipment, comprising: S1. The pulse current sensor is connected to the switchgear grounding conductor, and the ultrasonic sensor is attached to the switchgear insulation baffle to generate a synchronized current pulse sequence and ultrasonic waveform. S2. Based on the current pulse sequence and the ultrasonic waveform, the signal delay difference between the ground down conductor and the sensor at the insulating barrier is calculated; S3. Establish a multipath acoustic wave propagation database containing direct paths and expected reflection paths based on the three-dimensional structure of the switchgear metal cavity and the sensor layout coordinates; S4. Matching the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multipath database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance; S5. Combining the signal delay difference, the sensor layout coordinates and the equivalent propagation distance, an optimization algorithm is iterated to find the optimal spatial coordinates of the discharge point; S6. Mapping the optimal spatial coordinates to the switch cabinet structure topology diagram to generate a partial discharge location report for the electrical equipment.

[0006] Optionally, the pulse current sensor is connected to the grounding down conductor of the switch cabinet, and the ultrasonic sensor is attached to the insulating baffle of the switch cabinet to generate a synchronized current pulse sequence and ultrasonic waveform, including: The high-frequency current signal in the grounding down conductor is collected by a pulse current sensor and a standardized current pulse sequence is generated after filtering. The ultrasonic sensor receives the sound wave signal inside the switch cabinet and generates an ultrasonic waveform after noise suppression processing; A time synchronization device is used to perform clock alignment on the current pulse sequence and the ultrasonic waveform.

[0007] Optionally, the calculating, based on the current pulse sequence and the ultrasonic waveform, a signal delay difference between the ground down conductor and a sensor at the insulating barrier includes: Extracting the first pulse peak moment of the current pulse sequence as a reference moment; Identifying a starting fluctuation point in the ultrasonic waveform corresponding to the reference time; The time difference between the reference time and the starting fluctuation point is calculated as the initial delay difference.

[0008] Optionally, establishing a multi-path database of acoustic wave propagation including direct paths and expected reflection paths based on the three-dimensional structure of the metal cavity of the switch cabinet and the sensor layout coordinates includes: Measure the inner wall dimensions, corner angles, and insulation baffle positions of the switchgear metal cavity to build a three-dimensional structural model; Marking the layout coordinates of the pulse current sensor and the ultrasonic sensor in the three-dimensional structural model to determine the straight-line distance between the sensors; Simulating a direct path of an acoustic wave from a discharge point to the ultrasonic sensor, and recording the path length and propagation medium properties; Simulate the expected reflection path of the sound wave after it is reflected by the metal cavity wall and the insulating baffle, and record the number of reflections, the coordinates of the reflection point, and the path length; The parameters of the direct path and the expected reflection path are summarized to establish a multi-path database for acoustic wave propagation.

[0009] Optionally, the direct path of the simulated sound wave from the discharge point to the ultrasonic sensor, and recording of the path length and propagation medium properties, includes: Setting a virtual discharge point set in the three-dimensional structural model, wherein the virtual discharge point set covers a potential discharge area of ​​the metal cavity of the switch cabinet; For each virtual discharge point, a straight line propagation trajectory from the virtual discharge point to the ultrasonic sensor is generated as a direct path, and the path length of the straight line propagation trajectory and the sound velocity characteristics of the propagation medium are recorded.

[0010] Optionally, matching the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multipath database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance includes: Extracting the peak amplitude, rising edge slope and waveform duration of the first arriving wave from the ultrasonic waveform as the arriving wave characteristic parameters; Calling the theoretical waveform characteristic parameters corresponding to each path in the acoustic wave propagation multi-path database; Calculate the similarity between the characteristic parameters of the first arrival wave and the characteristic parameters of each theoretical waveform; Based on the path corresponding to the theoretical waveform with the highest similarity, the actual sound wave propagation path type and equivalent propagation distance are determined.

[0011] Optionally, the iteratively determining the optimal spatial coordinates of the discharge point by using an optimization algorithm based on the signal delay difference, the sensor layout coordinates, and the equivalent propagation distance includes: Using the layout coordinates of the pulse current sensor and the ultrasonic sensor as the boundary points, a cubic space containing the line connecting the two and the surrounding area is defined as the initial spatial coordinate range of the discharge point; Determining a distance constraint condition between the discharge point and the two sensors based on the signal delay difference and the equivalent propagation distance; Based on the distance constraint, constructing an objective function with the spatial coordinates of the discharge point as variables, wherein the output value of the objective function is the sum of the deviations of the distance constraint; Iteratively calculating the objective function using a particle swarm optimization algorithm, and updating the candidate coordinates of the discharge point in each iteration; When the calculation error of the candidate coordinates is less than a preset threshold, the coordinates are output as the optimal spatial coordinates.

[0012] Optionally, the adopting a particle swarm optimization algorithm to iteratively calculate the objective function, and updating the candidate coordinates of the discharge point in each iteration, includes: Initialize the particle positions of the particle swarm, where each particle corresponds to a candidate coordinate of the discharge point, and the particle positions are distributed within the initial spatial coordinate range; Taking the output value of the objective function as fitness, updating the speed and position of the particles through the particle swarm optimization algorithm, and iteratively optimizing the candidate coordinates; Calculate the output value of the objective function corresponding to the candidate coordinates after each iteration, and use the output value as the calculation error.

[0013] Optionally, the signal delay difference is used to correct the equivalent propagation distance, specifically including: Calculating a time correction value for sound wave propagation based on the signal delay difference; Based on the sound velocity of the air medium in the switch cabinet, multiplying the time correction value by the sound velocity to obtain a distance correction value; The distance correction value is subtracted from the equivalent propagation distance to obtain a corrected actual propagation distance, wherein the corrected actual propagation distance is used as an input parameter for iterative solution of the optimization algorithm.

[0014] Optionally, mapping the optimal spatial coordinates to a switch cabinet structure topology diagram to generate a partial discharge location report for electrical equipment includes: Converting the optimal spatial coordinates into two-dimensional coordinates in a switch cabinet structure topology diagram; Mark the internal components of the switch cabinet corresponding to the coordinates, including busbars, insulators, and circuit breakers; Count the occurrence frequency and discharge intensity of discharge points within the preset time; A positioning report including positioning position, discharge intensity and potential fault type is generated in combination with component attributes of the internal components of the switch cabinet.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention significantly improves the accuracy and efficiency of partial discharge positioning by establishing a multi-path database for acoustic wave propagation and optimizing the delay correction architecture: First, a sound wave propagation database of direct and reflected paths is pre-built based on the three-dimensional structure of the switchgear. Combined with ultrasonic waveform feature matching technology, the actual propagation path type and equivalent propagation distance are automatically identified. This effectively decouples the signal aliasing problem caused by multipath reflections, avoids reliance on manual experience, and shortens path analysis time. Secondly, the equivalent propagation distance is dynamically corrected by signal delay differences. Combining sensor placement coordinates with a particle swarm optimization algorithm, a distance-constrained objective function is constructed and the optimal spatial coordinates of the discharge point are iteratively determined. This optimization architecture transforms delay errors into spatial constraints, reducing repeated iterative calculations and significantly reducing positioning time while improving the robustness of the coordinate solution. Finally, the discharge point is mapped to the structural topology and associated with the component properties to generate a positioning report, providing an accurate basis for rapid diagnosis of insulation defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of a method for locating and detecting partial discharge in electrical equipment according to an embodiment of the present invention; The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0017] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] The embodiment of the present application provides a method for locating and detecting partial discharge of electrical equipment. The execution subject of the method for locating and detecting partial discharge of electrical equipment includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for locating and detecting partial discharge of electrical equipment can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks, and big data and artificial intelligence platforms.

[0019] Reference Figure 1 FIG. 1 is a flow chart of a method for locating and detecting partial discharge of electrical equipment according to an embodiment of the present invention. In this embodiment, the method for locating and detecting partial discharge of electrical equipment includes: S1. The pulse current sensor is connected to the grounding down conductor of the switch cabinet, and the ultrasonic sensor is attached to the insulating baffle of the switch cabinet to generate a synchronized current pulse sequence and ultrasonic waveform.

[0020] In an embodiment of the present invention, the pulse current sensor is connected to the grounding down conductor of the switch cabinet, and the ultrasonic sensor is attached to the insulating baffle of the switch cabinet to generate a synchronized current pulse sequence and ultrasonic waveform, including: The high-frequency current signal in the grounding down conductor is collected by a pulse current sensor and a standardized current pulse sequence is generated after filtering. The ultrasonic sensor receives the sound wave signal inside the switch cabinet and generates an ultrasonic waveform after noise suppression processing; A time synchronization device is used to perform clock alignment on the current pulse sequence and the ultrasonic waveform.

[0021] Specifically, a pulse current sensor is a device that detects current changes and converts them into a measurable signal. In practical applications, to obtain current information in a ground down conductor, a pulse current sensor is attached to the ground down conductor of a switchgear. This operation is similar to connecting a monitoring node to a circuit. The sensor's internal induction coil or other sensing element senses the current flowing through the ground down conductor.

[0022] For example, in common power facility layouts, the grounding down conductor is a critical connection between the switchgear and the earth. When an abnormality such as an electrical fault occurs inside the switchgear, abnormal current flows through the grounding down conductor to the earth. At this time, a pulse current sensor attached to the grounding down conductor can capture these current changes.

[0023] Specifically, an ultrasonic sensor is a device used to detect acoustic signals. In this invention, it is attached to the insulating baffle of the switchgear. The insulating baffle is a component within the switchgear used to isolate electrical components and ensure electrical insulation. Attaching the ultrasonic sensor to it effectively receives acoustic signals generated by phenomena such as electrical discharge within the switchgear. For example, when partial discharge occurs within the switchgear, ultrasonic waves are generated. When attached to the insulating baffle, the ultrasonic sensor receives these ultrasonic signals and converts them into electrical signals for subsequent processing.

[0024] In detail, the high-frequency current signal in the ground down conductor is collected by the pulse current sensor, and a standardized current pulse sequence is generated after filtering, including: When a pulse current sensor is operating, its internal sensing mechanism senses the current in the ground down conductor. In actual power system operation, in addition to possible fault-related high-frequency current signals, the ground down conductor also contains various low-frequency interference currents and interference signals such as environmental noise. To obtain accurate fault-related high-frequency current signals, the collected raw current signal must be filtered. This filtering is typically accomplished using a filter. Based on its designed frequency characteristics, the filter allows signals within a specific frequency range to pass while blocking signals at other frequencies.

[0025] For example, a high-pass filter can be used, with its cutoff frequency set to filter out low-frequency interference signals and only allow high-frequency current signals to pass through. After such filtering, a relatively pure high-frequency current signal is obtained.

[0026] The filtered high-frequency current signal is then further processed to generate a standardized current pulse train. This process may involve operations such as signal amplitude adjustment and pulse shaping.

[0027] For example, a comparator circuit compares the filtered signal with a specific threshold. When the signal amplitude exceeds the threshold, a pulse signal with a standard amplitude is output. This series of processes ultimately forms a standardized current pulse sequence with a specific amplitude, width, and interval. This standardized current pulse sequence facilitates subsequent analysis and processing, providing more accurate data for determining the operating status of the switchgear.

[0028] Specifically, the ultrasonic sensor receives acoustic signals from within the switchgear and generates an ultrasonic waveform after noise suppression. This involves attaching the ultrasonic sensor to the switchgear's insulating baffle, where it receives various acoustic signals from within the switchgear. However, in a real-world operating environment, in addition to acoustic signals related to electrical faults within the switchgear, there is also a significant amount of ambient noise and interfering acoustic signals generated by other equipment. To obtain an acoustic signal that accurately reflects the electrical conditions within the switchgear, noise suppression is required on the original acoustic signal received by the ultrasonic sensor.

[0029] Noise suppression can be achieved through various techniques, such as adaptive filtering algorithms based on digital signal processing. These algorithms automatically adjust filter parameters based on the characteristics of the received signal to maximize noise suppression. In practical applications, a sample of data containing noise but not the acoustic signal of the switchgear internal fault is collected. The statistical characteristics of this sample data are analyzed to determine the initial parameters of the adaptive filter. Then, when the ultrasonic sensor receives a mixed signal containing the fault acoustic signal and noise, the adaptive filter continuously adjusts its parameters based on the real-time received signal to suppress the noise. After noise suppression, a relatively pure acoustic signal related to the electrical phenomena within the switchgear is obtained. This signal is further processed to generate an ultrasonic waveform. Generating the ultrasonic waveform may involve converting the signal into visual waveform data. For example, an analog-to-digital converter converts the analog acoustic signal into a digital signal. A graphics algorithm is then used to generate the corresponding waveform on a display screen or data storage device, allowing for intuitive observation and analysis of the acoustic signal's characteristics. These characteristics can provide important clues to determine the presence and type of fault within the switchgear.

[0030] In detail, the clock alignment of the current pulse sequence and the ultrasonic waveform using a time synchronization device includes: A time synchronization device is a device that provides a precise time reference and enables time synchronization between different devices or signals. In the present invention, since the current pulse sequence and ultrasonic waveform are collected and generated by different sensors, their temporal correspondence is crucial for accurately analyzing the operating status of the switchgear.

[0031] For example, when an electrical fault occurs inside a switchgear cabinet, the resulting current changes and acoustic wave signals are sequential and closely correlated. If the timing of the current pulse sequence and the ultrasonic waveform are out of sync, errors can occur in the subsequent analysis of the correlation between the two, leading to inaccurate fault diagnosis.

[0032] Therefore, a time synchronization device is used to align the clocks of the current pulse train and the ultrasonic waveform. This device can achieve synchronization through various methods, most commonly using the Global Positioning System (GPS). The GPS system provides a highly accurate time signal. After receiving the GPS signal, the time synchronization device calibrates its internal clock to the GPS time.

[0033] Then, the time synchronization device sends a synchronization clock signal to the pulse current sensor and the ultrasonic sensor. After receiving the synchronization clock signal, the pulse current sensor and the ultrasonic sensor use it as a reference to record the timestamp of the data collected by each.

[0034] This way, when the current pulse train and ultrasonic waveform are generated, the time information they carry is based on the same time reference, achieving temporal alignment between the two. This clock alignment ensures that subsequent combined analysis of the current pulse train and ultrasonic waveform accurately identifies electrical phenomena within the switchgear based on their temporal relationships, improving the accuracy and reliability of switchgear operating status monitoring and fault diagnosis.

[0035] S2. Based on the current pulse sequence and the ultrasonic waveform, calculate the signal delay difference between the ground down conductor and the sensor at the insulating barrier.

[0036] In an embodiment of the present invention, the calculating, based on the current pulse sequence and the ultrasonic waveform, the signal delay difference between the ground down conductor and the sensor at the insulating barrier includes: Extracting the first pulse peak moment of the current pulse sequence as a reference moment; Identifying a starting fluctuation point in the ultrasonic waveform corresponding to the reference time; The time difference between the reference time and the starting fluctuation point is calculated as the initial delay difference.

[0037] In detail, the calculating of the signal delay difference between the ground down conductor and the sensor at the insulating barrier based on the current pulse sequence and the ultrasonic waveform includes: Signal delay difference refers to the difference in time it takes for two different signals to reach a monitoring point during their propagation. In this invention, the current pulse train is generated by a pulse current sensor, and the ultrasonic waveform is generated by an ultrasonic sensor. Due to the different placement of the two sensors and differences in signal generation and propagation paths, the time it takes for the two signals to reach their respective sensors will differ. Calculating this delay difference provides a temporal reference for subsequently determining the location of the discharge point.

[0038] For example, when partial discharge occurs inside the switch cabinet, current signals and acoustic signals are generated simultaneously. The current signal propagates through the grounding down conductor, and the acoustic signal propagates in the air. The two arrive at the sensor at different times. By calculating the time delay difference, the location of the discharge point can be inferred by combining information such as the signal propagation speed.

[0039] Specifically, extracting the first pulse peak moment of the current pulse sequence as the reference moment includes the following steps: a current pulse sequence is a series of pulse signals, each pulse having varying amplitudes. The first pulse peak moment refers to the time point when the first pulse in the current pulse sequence reaches its maximum amplitude. This moment is extracted as the reference moment because when partial discharge occurs, the current signal typically forms a detectable pulse first, and the peak moment of the first pulse is clearly characteristic and can serve as a starting point for a time reference.

[0040] During the implementation process, the waveform analysis of the current pulse sequence is required. Digital signal processing technology can be used to scan the generated standardized current pulse sequence. For example, a data acquisition frequency is set, assuming that 1000 data points are collected per second, and a curve showing the amplitude of the current pulse sequence changing with time is presented. By traversing the data points, the time coordinate corresponding to the maximum amplitude of the first pulse in the sequence is found. This time coordinate is the time of the first pulse peak. For example, in the collected sequence, the 100th data point corresponds to the maximum amplitude of the first pulse. If each data point is separated by 1 millisecond, the time of the first pulse peak is 100 milliseconds, which is determined as the reference time.

[0041] In detail, the identifying the starting fluctuation point in the ultrasonic waveform corresponding to the reference time includes: The starting point of fluctuation is the point in the ultrasonic waveform where noticeable fluctuations begin, relative to the reference time. Because the acoustic wave signal generated by partial discharge takes time to propagate, the start of fluctuations in the ultrasonic waveform may occur later than the reference time. Identifying this point is crucial for determining when the acoustic wave signal reaches the ultrasonic sensor.

[0042] During implementation, the ultrasonic waveform needs to be analyzed in time correlation with the reference moment. Based on the clock alignment achieved by the time synchronization device, the time axis of the ultrasonic waveform is kept consistent with the time axis of the current pulse sequence. The ultrasonic waveform is monitored in real time. When the reference moment is determined, the corresponding position is found on the time axis of the ultrasonic waveform, and the waveform changes after that position are observed. When the waveform begins to show obvious amplitude fluctuations from a relatively stable state, the starting point of the fluctuation is the starting fluctuation point. For example, the reference moment is 100 milliseconds. In the ultrasonic waveform, after 100 milliseconds, starting from 102 milliseconds, the amplitude of the waveform begins to fluctuate up and down from the previous stable value, then 102 milliseconds is the corresponding starting fluctuation point.

[0043] The technical effect of this step is to determine the arrival time of the sound wave signal. Combined with the reference time, it provides another key time point for calculating the delay difference, solving the problem of not being able to accurately obtain the arrival time of the sound wave signal.

[0044] Specifically, calculating the time difference between the reference time and the starting fluctuation point as the initial delay difference includes: The time difference refers to the interval between two different time points. In the present invention, the time difference between the reference moment and the starting fluctuation point is calculated, and the result obtained is the initial delay difference, which reflects the time difference between the current signal and the acoustic wave signal reaching their respective sensors.

[0045] The implementation process is to subtract the determined base time from the time value of the starting fluctuation point. For example, if the base time is 100 milliseconds and the starting fluctuation point is 102 milliseconds, the time difference is 102 milliseconds minus 100 milliseconds, which equals 2 milliseconds. This 2 milliseconds is used as the initial delay difference.

[0046] This step obtains the time difference between the two signal propagations, providing key data for the subsequent calculation of distance in combination with the propagation speed, and solving the problem of being unable to infer the difference in propagation paths due to unknown time difference.

[0047] In summary, the process of calculating the time delay difference forms a complete process, from determining the reference moment to identifying the corresponding starting fluctuation point and then calculating the time difference. These steps accurately obtain the signal delay difference, providing important time parameters for subsequently establishing a database of acoustic wave propagation paths and determining the coordinates of the discharge point. This solves the problem of large positioning errors caused by inaccurate time information in traditional partial discharge location. Extracting the first pulse peak moment and identifying the corresponding starting fluctuation point improves the accuracy of the time delay difference calculation by providing a clear time reference and corresponding fluctuation point identification.

[0048] S3. Based on the three-dimensional structure of the switchgear metal cavity and the sensor layout coordinates, a multi-path database of acoustic wave propagation including direct paths and expected reflection paths is established.

[0049] In an embodiment of the present invention, establishing a multi-path database of acoustic wave propagation including direct paths and expected reflection paths based on the three-dimensional structure of the metal cavity of the switch cabinet and the sensor layout coordinates includes: Measure the inner wall dimensions, corner angles, and insulation baffle positions of the switchgear metal cavity to build a three-dimensional structural model; Marking the layout coordinates of the pulse current sensor and the ultrasonic sensor in the three-dimensional structural model to determine the straight-line distance between the sensors; Simulating a direct path of an acoustic wave from a discharge point to the ultrasonic sensor, and recording the path length and propagation medium properties; Simulate the expected reflection path of the sound wave after it is reflected by the metal cavity wall and the insulating baffle, and record the number of reflections, the coordinates of the reflection point, and the path length; The parameters of the direct path and the expected reflection path are summarized to establish a multi-path database for acoustic wave propagation.

[0050] Specifically, the direct path of the simulated sound wave from the discharge point to the ultrasonic sensor, and recording of the path length and propagation medium properties, includes: Setting a virtual discharge point set in the three-dimensional structural model, wherein the virtual discharge point set covers a potential discharge area of ​​the metal cavity of the switch cabinet; For each virtual discharge point, a straight line propagation trajectory from the virtual discharge point to the ultrasonic sensor is generated as a direct path, and the path length of the straight line propagation trajectory and the sound velocity characteristics of the propagation medium are recorded.

[0051] Specifically, the method of establishing a multi-path database of acoustic wave propagation including direct paths and expected reflection paths based on the three-dimensional structure of the metal cavity of the switch cabinet and the sensor layout coordinates includes: The acoustic wave propagation multipath database stores information about the possible propagation paths of acoustic waves within the metal cavity of a switchgear. A direct path is the path that the acoustic wave takes from the discharge point to the ultrasonic sensor. An expected reflection path is the path that the acoustic wave takes from the discharge point, after reflecting off the metal cavity walls or insulating baffles, to the ultrasonic sensor. This database is established to subsequently match the detected acoustic wave signatures to the corresponding propagation paths, providing a path basis for locating the discharge point.

[0052] Specifically, the measurement of the inner wall dimensions, corner angles, and insulating baffle positions of the metal cavity of the switch cabinet and the construction of a three-dimensional structural model include: The metal cavity of a switchgear cabinet is the enclosed or semi-enclosed metal space within the cabinet, where electrical components are installed and operated. Inner wall dimensions refer to the length, width, and height of each inner wall of the metal cavity. Corner angle refers to the angle between the two intersecting inner walls at each corner of the metal cavity. Insulation baffle position refers to the specific spatial location of the insulation baffle within the metal cavity. A 3D structural model is a digital representation of the switchgear metal cavity's three-dimensional structure.

[0053] During implementation, a laser rangefinder can be used to measure the inner wall dimensions of the switchgear metal cavity. For example, the length, width, and height of the cavity are measured, assuming the results are 5 meters in length, 3 meters in width, and 2 meters in height respectively. For corner angles, an angle measuring instrument is used for measurement, such as measuring a corner angle of 90 degrees. For the position of the insulating baffle, its position is determined by setting reference points on the inner wall of the cavity and measuring the distance from the edge of the baffle to each reference point. Assuming that the lower left corner of the bottom surface of the cavity is used as the reference point, the horizontal distance from one edge of the baffle to the reference point is measured to be 1 meter and the vertical distance is 0.5 meters. The measured data is input into 3D modeling software, such as AutoCAD or SolidWorks. The software will automatically construct a 3D structural model of the switchgear metal cavity based on the input size, angle, and position information. This model can accurately reflect the three-dimensional shape of the cavity and the relative positions of each component.

[0054] In detail, marking the layout coordinates of the pulse current sensor and the ultrasonic sensor in the three-dimensional structural model and determining the straight-line distance between the sensors includes: The placement coordinates refer to the spatial coordinates of the sensor within the 3D structural model. These coordinates are composed of values ​​in the X, Y, and Z dimensions and are used to accurately represent the sensor's position. The linear distance between sensors refers to the length of the straight line connecting the pulse current sensor and ultrasonic sensor in 3D space.

[0055] The implementation process involves finding corresponding points in the three-dimensional structural model based on the actual locations of the pulse current sensor attached to the grounding down conductor and the ultrasonic sensor attached to the insulating baffle. For example, in the three-dimensional model, the coordinates of the pulse current sensor are determined to be (0,0,0), and the coordinates of the ultrasonic sensor are determined to be (5,3,2). Then, using the formula for calculating the distance between two points in three-dimensional space, that is, when the coordinates of the two points are (X1, Y1, Z1) and (X2, Y2, Z2), the above coordinates are substituted into the Euclidean distance formula to calculate the straight-line distance between the two sensors. The calculated straight-line distance between the two sensors is 6.16 meters.

[0056] In general, the three-dimensional structural model constructed in the previous step provides a platform for marking the sensor layout coordinates in this step. Without the three-dimensional model, it is impossible to accurately mark the sensor coordinates and calculate the straight-line distance in the virtual space.

[0057] Specifically, the direct path of the simulated sound wave from the discharge point to the ultrasonic sensor, and recording of the path length and propagation medium properties, includes: The discharge point is the location inside the switchgear where partial discharge occurs. The direct path is the path that the sound wave takes from the discharge point to the ultrasonic sensor without any reflection. The path length refers to the spatial length of this direct path. The propagation medium properties refer to the characteristics of the medium through which the sound wave propagates. In this scenario, the propagation medium is primarily air. Its properties include air temperature, humidity, and the propagation speed of the sound wave in the medium.

[0058] During implementation, based on the constructed three-dimensional structural model, assume the presence of a discharge point within the model. For example, set the coordinates of a discharge point to (2,1,1). Then, draw a straight line from the discharge point to the ultrasonic sensor coordinates (5,3,2) within the model. This line represents the direct path for sound wave propagation. Use the distance measurement tool in the model to measure the length of this line, assuming the path length is 4 meters. Regarding the propagation medium properties, since sound waves propagate through the air within the switchgear, a temperature and humidity sensor can be placed within the switchgear to measure the current air temperature of 25 degrees Celsius and humidity of 50%. Based on acoustic principles, under these temperature and humidity conditions, the propagation speed of sound waves in air is approximately 346 meters per second. These medium properties are then recorded.

[0059] In detail, a virtual discharge point set is set in the three-dimensional structural model, and the virtual discharge point set covers the potential discharge area of ​​the metal cavity of the switch cabinet, including: A virtual discharge point set is a collection of multiple virtual points in a 3D structural model where discharge may occur. A potential discharge area is a region within the metal cavity of a switchgear that is prone to partial discharge. These are typically areas where electrical components are concentrated or insulation is susceptible to damage.

[0060] During implementation, potential discharge areas are first determined based on the switchgear structure and the distribution of electrical components. For example, in a switchgear, areas where discharge may occur include the busbar-insulator connection and the circuit breaker contacts. Multiple virtual discharge points are then set evenly or as needed within these areas in the 3D structural model. For example, 100 virtual discharge points are set within the potential discharge area, with coordinates of each point being (x1, y1, z1), (x2, y2, z2), and so on (x100, y100, z100). These points together constitute a virtual discharge point set that fully covers the potential discharge area.

[0061] Specifically, for each virtual discharge point, generating a straight line propagation trajectory from the virtual discharge point to the ultrasonic sensor as a direct path, and recording the path length of the straight line propagation trajectory and the sound velocity characteristics of the propagation medium include: The linear propagation trajectory refers to the trajectory formed by the straight line from the virtual discharge point to the ultrasonic sensor. The sound velocity characteristic refers to the propagation speed characteristic of the sound wave in the propagation medium, that is, the speed of the sound wave propagating in the medium.

[0062] During implementation, for each virtual discharge point in the set of virtual discharge points, for example, one with coordinates (1, 0.5, 0.5), a straight line is generated in the 3D structural model from that point to the ultrasonic sensor coordinates (5, 3, 2). This line represents the linear propagation trajectory of the direct path corresponding to that virtual discharge point. The model's measurement function is used to measure the length of this line, assuming a path length of 3 meters. As for the sound velocity characteristics of the propagation medium, since the propagation medium is air, combined with the previously measured air temperature and humidity, the sound velocity is determined to be 346 meters per second. This sound velocity characteristic and the path length are recorded together.

[0063] Specifically, simulating the expected reflection path of the sound wave after it is reflected by the metal cavity wall and the insulating baffle, and recording the number of reflections, the coordinates of the reflection points, and the path length, includes: The expected reflection path is the path a sound wave takes after reflecting off a metal cavity wall or insulating baffle, simulated based on the laws of sound wave propagation, to reach the ultrasonic sensor. The number of reflections refers to the number of times a sound wave is reflected by a wall or baffle during propagation. The reflection point coordinates are the spatial coordinates of the point where the sound wave is reflected from the wall or baffle.

[0064] During implementation, the reflection path is simulated based on a three-dimensional structural model and the law of acoustic wave reflection, which states that the angle of incidence equals the angle of reflection. For example, for a virtual discharge point (2,1,1), the simulated sound wave first propagates to a wall of the metal cavity, assuming this wall is the right wall (at X = 5 meters). The reflection point at this wall is calculated. According to the law of reflection, the coordinates of the reflection point are calculated and assumed to be (5,2,1.5). The sound wave then propagates from this reflection point to the ultrasonic sensor (5,3,2). This path is the expected reflection path for a single reflection. The number of reflections is recorded as 1, the coordinates of the reflection point (5,2,1.5), and the total length of the path is measured, assuming it is 5 meters. For another example, the simulated sound wave path is recorded as first reflecting from the left wall, then from the insulating baffle, and finally reaching the sensor. The number of reflections is recorded as 2, along with the corresponding reflection point coordinates and path length.

[0065] In detail, the steps of summarizing the parameters of the direct path and the expected reflection path to establish a multi-path database for acoustic wave propagation include: Parameter aggregation is the process of collecting, organizing, and centrally storing the parameters of the direct path and the expected reflection path. The acoustic wave propagation multipath database is a structured data set that stores these aggregated parameters.

[0066] During implementation, design the database structure and set fields such as path type (direct or reflected), path length, propagation medium properties, number of reflections, and reflection point coordinates. Then, enter the parameters of the direct path and the expected reflection path recorded in the previous step according to the database structure requirements. For example, for a direct path, enter the path type as direct, path length as 4 meters, and propagation medium sound speed as 346 meters per second. For a reflection path, enter the path type as reflection, number of reflections as 1, reflection point coordinates (5, 2, 1.5), and path length as 5 meters. Integrate all this information into the database to form a complete multipath database for acoustic wave propagation.

[0067] Overall, the process from constructing a 3D model to establishing a multipath database provides comprehensive fundamental data for identifying acoustic wave propagation paths. Simulating multiple paths and summarizing their parameters to build a database breaks through the traditional limitation of considering only direct paths. By comprehensively considering possible reflection paths, the system enables more accurate determination of acoustic wave propagation paths, thereby improving the accuracy of partial discharge location.

[0068] S4. Match the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multi-path database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance.

[0069] In an embodiment of the present invention, matching the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multipath database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance includes: Extracting the peak amplitude, rising edge slope and waveform duration of the first arriving wave from the ultrasonic waveform as the arriving wave characteristic parameters; Calling the theoretical waveform characteristic parameters corresponding to each path in the acoustic wave propagation multi-path database; Calculate the similarity between the characteristic parameters of the first arrival wave and the characteristic parameters of each theoretical waveform; Based on the path corresponding to the theoretical waveform with the highest similarity, the actual sound wave propagation path type and equivalent propagation distance are determined.

[0070] Specifically, matching the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multipath database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance includes: The first-arrival wave signature refers to the characteristics of the first sound wave in the ultrasonic waveform to reach the ultrasonic sensor. These characteristics can reflect the propagation path of the sound wave. The actual sound wave propagation path type refers to whether the path the sound wave takes from the discharge point to the ultrasonic sensor is a direct path or a reflected path. The equivalent propagation distance refers to the length of the actual sound wave propagation path and is an important parameter for subsequent calculation of the discharge point location. Through matching, the actual detected sound wave signature can be correlated with the path characteristics preset in the database to determine the actual propagation path and corresponding distance.

[0071] Specifically, extracting the peak amplitude, rising edge slope, and waveform duration of the first arriving wave from the ultrasonic waveform as the arrival wave characteristic parameters includes: Peak amplitude refers to the maximum amplitude in the waveform of the first arrival wave, reflecting the energy of the sound wave at the time it reaches the sensor. Rising slope refers to the rate at which the amplitude of the first arrival wave changes over time from the initial fluctuation point to the peak amplitude. Waveform duration refers to the time from the initial fluctuation point of the first arrival wave to the return to a stable state. Arrival wave characteristic parameters are specific values ​​used to describe the characteristics of the first arrival wave.

[0072] During implementation, the ultrasonic waveform is analyzed with the help of signal processing software. For example, in the acquired ultrasonic waveform, the waveform of the first arrival wave shows obvious fluctuations. For the peak amplitude, the software scans the waveform data to find the maximum amplitude, assuming that the peak amplitude is 5 millivolts. For the rising edge slope, the waveform from the starting fluctuation point to the peak amplitude point is selected, and the ratio of the amplitude change to the time change of the waveform is calculated. Assuming that the amplitude rises from 0 to 5 millivolts in 2 milliseconds, the rising edge slope is 2.5 millivolts / millisecond. For the waveform duration, the starting fluctuation point and the ending point of the first arrival wave are determined, and the time difference between the two points is calculated. Assuming that it starts at 102 milliseconds and ends at 110 milliseconds, the waveform duration is 8 milliseconds. These three parameters are extracted as the characteristic parameters of the arrival wave.

[0073] In detail, the calling of the theoretical waveform characteristic parameters corresponding to each path in the acoustic wave propagation multi-path database includes: Theoretical waveform characteristic parameters refer to the characteristic parameters of the acoustic waveform corresponding to each path obtained through theoretical calculation or simulation based on the parameters of each path in the acoustic wave propagation multi-path database, including the theoretical peak amplitude, rising edge slope and waveform duration.

[0074] During implementation, a database query command is used to retrieve the theoretical waveform characteristic parameters of each path from an established database of multipath acoustic wave propagation. For example, the database stores theoretical parameters for one direct path and two reflected paths: the direct path's theoretical peak amplitude is 6 mV, its rising edge slope is 3 mV / ms, and its waveform duration is 7 msec; the first reflected path's theoretical peak amplitude is 4 mV, its rising edge slope is 2 mV / ms, and its waveform duration is 9 msec; and the second reflected path's theoretical peak amplitude is 3 mV, its rising edge slope is 1.5 mV / ms, and its waveform duration is 10 msec. These parameters are retrieved one by one and used in the subsequent similarity calculation.

[0075] Specifically, the calculation of the similarity between the characteristic parameters of the first arrival wave and the characteristic parameters of each theoretical waveform includes: Similarity refers to the degree of similarity between the characteristic parameters of the first arriving wave and the characteristic parameters of each theoretical waveform. The higher the similarity, the closer the actual sound wave is to the sound wave characteristics corresponding to the theoretical path.

[0076] During implementation, the similarity is calculated using a characteristic parameter comparison method. The differences between the actual arrival wave characteristic parameters and the corresponding parameters in the theoretical waveform characteristic parameters for each theoretical path are calculated separately, and the similarity is then calculated by combining these differences. For example, the differences in peak amplitude, rising edge slope, and waveform duration are used as the calculation basis, with each parameter given equal weight. The actual parameters are: peak amplitude 5 mV, rising edge slope 2.5 mV / ms, and waveform duration 8 ms. For the theoretical parameters of the direct path, the differences in peak amplitude are 1 mV, rising edge slope 0.5 mV / ms, and waveform duration 1 ms, for a total difference of 2.5. For the theoretical parameters of the first reflected path, the differences in peak amplitude are 1 mV, rising edge slope 0.5 mV / ms, and waveform duration 1 ms, for a total difference of 2.5. For the theoretical parameters of the second reflected path, the differences in peak amplitude are 2 mV, rising edge slope 1 mV / ms, and waveform duration 2 ms, for a total difference of 5. The smaller the total difference, the higher the similarity, so the actual parameters are more similar to the theoretical parameters of the first two paths than to the third path.

[0077] Specifically, determining the actual sound wave propagation path type and equivalent propagation distance based on the path corresponding to the theoretical waveform with the highest similarity includes: The actual sound wave propagation path type refers to whether the propagation path of the actual sound wave, determined based on the matching results, is a direct path or a reflected path. The equivalent propagation distance refers to the length of the path corresponding to the theoretical waveform with the highest similarity.

[0078] During implementation, the calculated similarities of each path are compared to identify the path corresponding to the theoretical waveform with the highest similarity. For example, if the actual arrival wave characteristic parameters differ by a total of 2.5 from the theoretical parameters of the direct path and also differ by a total of 2.5 from the theoretical parameters of the first reflected path, further analysis can be performed to determine the degree of matching between the parameters. If the differences in peak amplitude and rising edge slope are smaller, then one of the paths with the highest similarity can be determined. Assuming the direct path is ultimately determined to be the path with the highest similarity, then the actual sound wave propagation path type is a direct path. The path length recorded in the database for this direct path is 4 meters, so the equivalent propagation distance is 4 meters.

[0079] In summary, this step achieves accurate identification of the actual sound wave propagation path through a series of operations: extracting characteristic parameters, calling theoretical parameters, calculating similarity, and determining the actual path. This involves matching the actual characteristic parameters with the theoretical parameters in the database for similarity, and through a comprehensive multi-parameter comparison, improving the accuracy of path identification. This provides critical path information for subsequent precise positioning of the discharge point, forming a complete path identification process.

[0080] S5. Combining the signal delay difference, the sensor layout coordinates and the equivalent propagation distance, the optimal spatial coordinates of the discharge point are iteratively calculated through an optimization algorithm.

[0081] In an embodiment of the present invention, the iterative determination of the optimal spatial coordinates of the discharge point by using an optimization algorithm based on the signal delay difference, the sensor layout coordinates, and the equivalent propagation distance includes: Using the layout coordinates of the pulse current sensor and the ultrasonic sensor as the boundary points, a cubic space containing the line connecting the two and the surrounding area is defined as the initial spatial coordinate range of the discharge point; Determining a distance constraint condition between the discharge point and the two sensors based on the signal delay difference and the equivalent propagation distance; Based on the distance constraint, constructing an objective function with the spatial coordinates of the discharge point as variables, wherein the output value of the objective function is the sum of the deviations of the distance constraint; Iteratively calculating the objective function using a particle swarm optimization algorithm, and updating the candidate coordinates of the discharge point in each iteration; When the calculation error of the candidate coordinates is less than a preset threshold, the coordinates are output as the optimal spatial coordinates.

[0082] In detail, the particle swarm optimization algorithm is used to iteratively calculate the objective function, and each iteration updates the candidate coordinates of the discharge point, including: Initialize the particle positions of the particle swarm, where each particle corresponds to a candidate coordinate of the discharge point, and the particle positions are distributed within the initial spatial coordinate range; Taking the output value of the objective function as fitness, updating the speed and position of the particles through the particle swarm optimization algorithm, and iteratively optimizing the candidate coordinates; Calculate the output value of the objective function corresponding to the candidate coordinates after each iteration, and use the output value as the calculation error.

[0083] Specifically, the signal delay difference is used to correct the equivalent propagation distance, specifically including: Calculating a time correction value for sound wave propagation based on the signal delay difference; Based on the sound velocity of the air medium in the switch cabinet, multiplying the time correction value by the sound velocity to obtain a distance correction value; The distance correction value is subtracted from the equivalent propagation distance to obtain a corrected actual propagation distance, wherein the corrected actual propagation distance is used as an input parameter for iterative solution of the optimization algorithm.

[0084] In detail, the iterative calculation of the optimal spatial coordinates of the discharge point by combining the signal delay difference, the sensor layout coordinates, and the equivalent propagation distance through an optimization algorithm includes: The optimal spatial coordinates are the three-dimensional coordinates that most accurately represent the location of the discharge point within the metal cavity of the switchgear. The signal delay difference is the time difference between the current signal and the acoustic signal reaching the sensor. The sensor layout coordinates are the specific locations of the pulse current sensor and ultrasonic sensor in three-dimensional space. The equivalent propagation distance is the actual length of the acoustic wave propagation path. The solution is iteratively solved using an optimization algorithm, which uses mathematical calculations to continuously adjust the coordinate assumptions of the discharge point to ultimately find the coordinates that best match the actual situation.

[0085] Specifically, the layout coordinates of the pulse current sensor and the ultrasonic sensor are used as boundary points to define a cubic space including the line connecting the two and the surrounding area as the initial spatial coordinate range of the discharge point, including: Boundary points refer to the coordinates of the pulse current sensor and ultrasonic sensor in three-dimensional space. They serve as reference points for defining the initial spatial range. The cubic space is the three-dimensional space formed by the two sensor coordinates as boundary points, including the line connecting them and the surrounding area. The initial spatial coordinate range is a preliminary definition of the possible spatial range of the discharge point. This range provides the search boundary for subsequent iterative calculations.

[0086] During implementation, the pulse current sensor's layout coordinates are known to be (0,0,0) and the ultrasonic sensor's layout coordinates are known to be (5,3,2). Based on these two coordinates, a cube is defined in three-dimensional space. The side length of the cube can be set based on the actual dimensions of the switchgear. For example, based on the connection between the two sensors and extending 1 meter in each direction, the cube's x-axis range is -1 to 6, the y-axis range is -1 to 4, and the z-axis range is -1 to 3. This cube encompasses the connection between the two sensors and the surrounding area. Using this as the initial spatial coordinate range of the discharge point means that the coordinates of the discharge point are likely to fall within this range.

[0087] In detail, the determining of the distance constraint between the discharge point and the two sensors based on the signal delay difference and the equivalent propagation distance includes: Distance constraints are conditions that must be met between the discharge point and the pulse current sensor and ultrasonic sensor, derived from signal delay differences and equivalent propagation distances. These conditions, derived from the propagation laws of sound waves and current, are used to limit the possible locations of the discharge point.

[0088] During implementation, assume a signal delay difference of 2 milliseconds and an equivalent propagation distance of 4 meters. Since current signals propagate extremely quickly, the propagation time from the discharge point to the pulse current sensor can be considered zero. Therefore, the distance from the discharge point to the pulse current sensor can be derived using other relationships. The distance from the discharge point to the ultrasonic sensor is an equivalent propagation distance of 4 meters. Furthermore, based on the signal delay difference, the time it takes for the sound wave to propagate 4 meters should be equal to the signal delay difference plus the current signal propagation time (approximately zero). Combined with the speed of sound of 346 meters per second, the rationality of 4 meters and 2 milliseconds can be verified (346 meters per second × 0.002 seconds ≈ 0.69 meters. Due to reflections and other factors, the actual equivalent propagation distance needs to be corrected, and this is only an example logic). The distance constraints are thus determined as follows: the distance from the discharge point to the ultrasonic sensor is approximately 4 meters, and the distance to the pulse current sensor must satisfy the time relationship with the sound wave propagation.

[0089] In detail, based on the distance constraint condition, an objective function with the spatial coordinates of the discharge point as a variable is constructed, and the output value of the objective function is the sum of the deviations of the distance constraint, including: The objective function is a mathematical function used to measure the degree to which the assumed discharge point coordinates conform to the distance constraints. The spatial coordinates of the discharge point are variables, meaning the function's input is the three-dimensional coordinates (x, y, z) of the discharge point. The function's calculation results change as the coordinates change. The sum of the distance constraint deviations is the sum of the differences between the assumed discharge point coordinates and each distance constraint. A smaller sum indicates a more accurate fit between the coordinate assumptions and the actual situation.

[0090] When implementing, assume that the spatial coordinates of the discharge point are (x, y, z), the pulse current sensor coordinates are (0, 0, 0), and the ultrasonic sensor coordinates are (5, 3, 2). According to the distance formula, the distance from the discharge point to the ultrasonic sensor is , this distance should be equal to the equivalent propagation distance of 4 meters, and its deviation is .

[0091] At the same time, the distance constraint from the discharge point to the pulse current sensor is determined by combining factors such as signal delay difference. Assuming it is a certain value, its deviation can also be expressed similarly. The objective function is the sum of these deviations, that is, , where L is the deviation of the other distance constraints.

[0092] This step transforms the problem of solving the coordinates of the discharge point into an optimization problem of a mathematical function, solving the problem of being unable to systematically solve the coordinates due to the lack of a mathematical model. The distance constraint is the basis for constructing the objective function, and the objective function is the mathematical expression of the distance constraint. Without the distance constraint, the objective function cannot be constructed.

[0093] In detail, the particle swarm optimization algorithm is used to iteratively calculate the objective function, and each iteration updates the candidate coordinates of the discharge point, including: The particle swarm optimization algorithm (PSO) is an optimization algorithm based on swarm intelligence. It simulates the foraging behavior of a flock of birds, allowing multiple candidate solutions (particles) to continuously move through the search space to find the optimal solution. Candidate coordinates refer to the possible coordinates of a hypothetical discharge point during the iterative calculation process, and these coordinates are continuously updated with each iteration. Iterative calculation involves repeated calculations, each time adjusting the candidate coordinates based on the previous results, gradually approaching the optimal solution.

[0094] During implementation, a particle swarm is initialized within the initial spatial coordinate range. Suppose there are 50 particles, each representing a candidate coordinate for a discharge point. These coordinates are randomly distributed within the cubic space. For example, the initial candidate coordinates for a particle are (2, 1, 1). Each candidate coordinate is substituted into the objective function, and the sum of deviations is calculated. According to the rules of the particle swarm optimization algorithm, each particle adjusts its movement direction and distance based on its own historical optimal position and the global optimal position of the entire particle swarm. For example, if the sum of deviations corresponding to a particle's current candidate coordinates is 3, while its historical optimal sum of deviations is 2, and the global optimal sum of deviations of the particle swarm is 1, then the particle will move toward the global optimal position, and the updated candidate coordinates may be (2.5, 1.2, 1.1). Each iteration updates the candidate coordinates of all particles, and this process repeats.

[0095] Specifically, the particle positions of the initialized particle swarm, each particle corresponding to a candidate coordinate of the discharge point, and the particle positions distributed within the initial spatial coordinate range, include: The particle position refers to the specific coordinate value of the particle within the initial spatial coordinate range. Each particle position corresponds to a hypothetical candidate coordinate of the discharge point. Initialization refers to assigning an initial position to each particle in the particle swarm before the iterative calculation begins.

[0096] During implementation, a random number generator is used to generate coordinates for each particle based on the x, y, and z axis ranges of the initial spatial coordinate range. For example, if the initial spatial x-axis range is -1 to 6, the y-axis range is -1 to 4, and the z-axis range is -1 to 3, then for 50 particles, each particle's x-coordinate is randomly generated between -1 and 6, its y-coordinate is randomly generated between -1 and 4, and its z-coordinate is randomly generated between -1 and 3, ensuring that all particle positions are distributed within the initial spatial coordinate range. For example, if one particle's initial position is (1, 0.5, 0.5), this position is a candidate coordinate for a discharge point.

[0097] This step provides the initial search starting point for the particle swarm optimization algorithm, solving the problem of being unable to start iterative calculations due to the lack of initial candidate coordinates. The previous step determines the initial spatial coordinate range, which provides a spatial boundary for initializing the particle position in this step. Particle positions can only be generated within this range.

[0098] In detail, the output value of the objective function is used as fitness, the speed and position of the particles are updated by the particle swarm optimization algorithm, and the candidate coordinates are iteratively optimized, including: Fitness is a metric used to evaluate the quality of a particle's position. In this method, the output value of the objective function (the sum of deviations) is the fitness. A smaller fitness indicates that the candidate coordinates corresponding to the particle position are closer to the optimal solution. Particle velocity refers to the rate and direction of movement of the particle in three-dimensional space, which determines the magnitude and direction of the particle position update. Iterative optimization involves continuously updating the particle's velocity and position to gradually move the candidate coordinates closer to the optimal solution.

[0099] During implementation, each particle's fitness is the output value of the objective function corresponding to that particle. For example, if a particle's fitness is 2.5 and another's is 1.8, the particle with a fitness of 1.8 has a better position. According to the particle swarm optimization algorithm, the particle speed update formula can be expressed as: New Speed ​​= Inertia Factor × Current Speed ​​+ Individual Learning Factor × Random Number × (Individual Optimal Position - Current Position) + Group Learning Factor × Random Number × (Group Optimal Position - Current Position).

[0100] Assume that the inertia factor is 0.5, the individual learning factor is 1, the group learning factor is 1, the current velocity of a particle is (0.1, 0.1, 0.1), the difference between the individual optimal position and the current position is (0.5, 0.3, 0.2), the difference between the group optimal position and the current position is (1, 0.8, 0.5), and the random numbers are all 0.5, then the new velocity v=0.5×(0.1, 0.1, 0.1)+1×0.5×(0.5, 0.3, 0.2)+1×0.5×(1, 0.8, 0.5)=(0.05+0.25+0.5,0.05+0.15+0.4,0.05+0.1+0.25)=(0.8, 0.6, 0.4). Update the particle position according to the new velocity. The new position = current position + new velocity. Assuming the current position is (2, 1, 1), the new position is (2.8, 1.6, 1.4), which is the optimized candidate coordinate.

[0101] This step guides particles toward the optimal solution through scientific speed and position update rules, solves the problem of lack of basis for updating candidate coordinates, and improves the efficiency of iterative optimization.

[0102] Specifically, calculating the output value of the objective function corresponding to the candidate coordinates after each iteration and using the output value as the calculation error includes: The calculation error refers to the output value (total deviation) of the objective function corresponding to the candidate coordinates after each iteration, which reflects the degree of deviation between the current candidate coordinates and the actual discharge point coordinates.

[0103] When implementing, after each iteration, the candidate coordinates of the particle are updated, and the new candidate coordinates are substituted into the objective function for calculation. For example, after the iteration, the candidate coordinates of a particle are (3, 2, 1.5), and they are substituted into the objective function , the calculated output value is 1.2, which is the calculation error corresponding to the candidate coordinate.

[0104] This step provides a basis for judging whether the iteration has converged, and solves the problem of not being able to determine when to stop the iteration due to the inability to measure the quality of the candidate coordinates.

[0105] Specifically, when the calculation error of the candidate coordinates is less than a preset threshold, outputting the coordinates as the optimal spatial coordinates includes: The preset threshold is a pre-set error limit. When the calculated error is less than this limit, it means that the candidate coordinates are close enough to the actual discharge point coordinates. The optimal spatial coordinates are the candidate coordinates whose calculated error is less than the preset threshold. They are the final result of the iterative calculation.

[0106] During implementation, the preset threshold can be set according to the positioning accuracy requirements, for example, set to 0.5. During the iterative calculation process, the calculation error of each candidate coordinate is continuously monitored. When the calculation error of a candidate coordinate is 0.4, which is less than the preset threshold of 0.5, it means that the coordinate has met the accuracy requirements. The coordinate (for example, (3.2, 2.1, 1.6)) is output as the optimal spatial coordinate of the discharge point.

[0107] In detail, the time correction value of sound wave propagation calculated based on the signal delay difference includes: the time correction value refers to the numerical value for adjusting the sound wave propagation time based on the signal delay difference, which is used to correct the calculation deviation of the sound wave propagation time caused by various factors.

[0108] During implementation, the signal delay difference is 2 milliseconds. Since the current signal propagation time is extremely short and can be ignored, the actual time of sound wave propagation should be the signal delay difference plus other possible time deviations. Assume that after analysis, the time correction value is 0.5 milliseconds (this is only an example and the actual calculation needs to be based on the specific situation).

[0109] In detail, the method of multiplying the time correction value by the sound velocity of the air medium in the switch cabinet to obtain the distance correction value includes: The distance correction value is calculated by combining the time correction value and the speed of sound, and is used to correct the equivalent propagation distance. The speed of sound in air is the speed at which sound waves propagate in the air inside the switch cabinet, and is set to 346 m / s in this invention.

[0110] During implementation, it is known that the time correction value is 0.5 milliseconds (i.e. 0.0005 seconds) and the speed of sound is 346 meters per second, then the distance correction value = time correction value × speed of sound = 0.0005 × 346 = 0.173 meters.

[0111] In detail, the equivalent propagation distance is subtracted from the distance correction value to obtain a corrected actual propagation distance, wherein the corrected actual propagation distance is used as an input parameter for iterative solution of the optimization algorithm, including: The corrected actual propagation distance is the equivalent propagation distance after adjusting for the distance correction value, which is closer to the actual distance the sound wave travels. Input parameters are the data required during the iterative solution of the optimization algorithm. The corrected actual propagation distance is used to adjust distance constraints, etc.

[0112] In practice, the equivalent propagation distance is 4 meters, and the distance correction value is 0.173 meters. Therefore, the corrected actual propagation distance = 4 - 0.173 = 3.827 meters. This distance is used as an input parameter in the optimization algorithm to update the distance constraints. For example, the distance constraint from the discharge point to the ultrasonic sensor is adjusted to 3.827 meters.

[0113] The correction steps improve the accuracy of the equivalent propagation distance and solve the problem of reduced positioning accuracy due to deviation in propagation distance calculation.

[0114] In general, this step involves a series of operations, including defining a range, establishing constraints, and iterating optimizations, ultimately determining the optimal spatial coordinates of the discharge point. The particle swarm optimization algorithm is used for iterative solving, combined with distance constraints and a correction mechanism, to efficiently and accurately locate the discharge point coordinates, overcoming the low precision and inefficiency of traditional positioning methods.

[0115] S6. Mapping the optimal spatial coordinates to the switch cabinet structure topology diagram to generate a partial discharge location report for the electrical equipment.

[0116] In an embodiment of the present invention, mapping the optimal spatial coordinates to a switchgear structural topology diagram to generate a partial discharge location report for electrical equipment includes: Converting the optimal spatial coordinates into two-dimensional coordinates in a switch cabinet structure topology diagram; Mark the internal components of the switch cabinet corresponding to the coordinates, including busbars, insulators, and circuit breakers; Count the occurrence frequency and discharge intensity of discharge points within the preset time; A positioning report including positioning position, discharge intensity and potential fault type is generated in combination with component attributes of the internal components of the switch cabinet.

[0117] Specifically, mapping the optimal spatial coordinates to a switchgear structural topology diagram to generate a partial discharge location report for electrical equipment includes: The optimal spatial coordinates refer to the three-dimensional coordinates of the discharge point, obtained through iterative optimization algorithms. A switchgear topology diagram is a two-dimensional schematic diagram that illustrates the layout and connectivity of components within the switchgear. A partial discharge location report contains the location of the discharge point, relevant component information, and potential faults. Mapping is the process of converting the three-dimensional optimal spatial coordinates into a two-dimensional topology diagram. Generating a report visually presents information related to partial discharge, providing a basis for equipment maintenance.

[0118] In detail, converting the optimal spatial coordinates into two-dimensional coordinates in the switch cabinet structure topology diagram includes: Two-dimensional coordinates are used to represent positions on the plane of a switchgear topology diagram. They typically consist of values ​​in both the horizontal and vertical dimensions. Conversion is the process of mapping optimal three-dimensional spatial coordinates onto a two-dimensional topology diagram using specific rules.

[0119] When implementing, the optimal spatial coordinates of the known discharge point are (3.2, 2.1, 1.6). The switchgear structure topology diagram uses the bottom surface of the switchgear as the projection surface to establish a two-dimensional coordinate system, with the horizontal axis as the X axis and the vertical axis as the Y axis. When converting, ignore the Z axis coordinate (or correspond to different areas of the topology diagram according to the height), and directly correspond the X and Y values ​​of the three-dimensional coordinates to the X and Y axes of the topology diagram, for example (3.2, 2.1). After the ratio conversion with the topology diagram (assuming the topology Figure 1 Centimeters represent actual 0.1 meters), and the two-dimensional coordinates on the topological map are (32 cm, 21 cm).

[0120] Specifically, the internal components of the switch cabinet corresponding to the marked coordinates include busbars, insulators, and circuit breakers, including: Markings are two-dimensional coordinates placed on the switchgear topology diagram to indicate the location of the discharge point. Switchgear internal components refer to the electrical components installed within the switchgear. Busbars are conductors used to transmit electrical energy, insulators are insulating components, and circuit breakers are devices used to open or close circuits.

[0121] During implementation, the position corresponding to the two-dimensional coordinates (32 cm, 21 cm) is found on the topological map. By checking the component layout corresponding to the position, it is determined that the position belongs to the area near the busbar. Then the "discharge point" is marked at the coordinates on the topological map, and it is noted that the corresponding component is the busbar.

[0122] This step clarifies the components associated with the discharge point, solving the problem of being unable to perform targeted maintenance due to not knowing the corresponding components of the discharge point.

[0123] Specifically, the statistical generation of the frequency and intensity of discharge points within a preset time period includes: The preset time is a pre-set period used to count discharge events, such as one hour. The frequency of occurrence refers to the number of times partial discharges occur at a discharge point within the preset time. The discharge intensity refers to the strength of the partial discharge, typically measured by parameters such as the amplitude of the current pulse train. Statistics are the process of recording and calculating both frequency and intensity.

[0124] During implementation, a preset time of one hour is set, and the monitoring equipment records the discharge situation at the discharge point during this time period. Suppose that the monitoring device detects five partial discharges at the discharge point within one hour. By analyzing the peak amplitude of the current pulse sequence, it is determined that the discharge intensity is medium, and these data are statistically analyzed.

[0125] This step obtains information on the activity patterns and strength of the discharge point, resolving the problem of being unable to assess the severity of the fault due to a lack of discharge frequency and intensity data. The previous step determined the location of the discharge point, and this step conducts targeted statistics based on that location. Without a clear location, the statistics lose their purpose.

[0126] In detail, the generating of a positioning report including positioning position, discharge intensity and potential fault type in combination with the component attributes of the internal components of the switch cabinet includes: Component attributes refer to the characteristics of switchgear components, such as busbar material and insulator insulation rating. Positioning refers to the location of the discharge point on the switchgear topology diagram. Potential fault types, such as insulation aging and poor contact, are inferred based on the discharge point location, component attributes, and discharge intensity. Report generation is the process of organizing this information into standardized documentation.

[0127] During implementation, it was determined that the discharge point was located near the busbar, the discharge intensity was medium, and the busbar components were copper with a rated current of 500A. Based on this information, the potential fault type was suspected to be partial discharge caused by damaged busbar insulation. The location (two-dimensional coordinates and corresponding components on the topology map), the discharge intensity (medium), and the potential fault type (damaged busbar insulation) were documented to create a partial discharge location report.

[0128] In summary, this step completes the final presentation of partial discharge location through coordinate transformation, component labeling, data statistics, and report generation. Inferring potential fault types based on component attributes involves correlating discharge information with component characteristics, enhancing the practical value of the report and providing precise guidance for equipment maintenance.

[0129] In the several embodiments provided by the present invention, it should be understood that the disclosed methods can be implemented in other ways.

[0130] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0131] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method and technology of using digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to achieve optimal results.

[0132] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for locating and detecting partial discharge of electrical equipment, characterized in that: The method comprises: S1. The pulse current sensor is connected to the switchgear grounding conductor, and the ultrasonic sensor is attached to the switchgear insulation baffle to generate a synchronized current pulse sequence and ultrasonic waveform. S2. Based on the current pulse sequence and the ultrasonic waveform, the signal delay difference between the ground down conductor and the sensor at the insulating barrier is calculated; S3. Establish a multipath acoustic wave propagation database containing direct paths and expected reflection paths based on the three-dimensional structure of the switchgear metal cavity and the sensor layout coordinates; S4. Matching the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multipath database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance; S5. Combining the signal delay difference, the sensor layout coordinates and the equivalent propagation distance, an optimization algorithm is iterated to find the optimal spatial coordinates of the discharge point; S6. Mapping the optimal spatial coordinates to the switch cabinet structure topology diagram to generate a partial discharge location report for the electrical equipment.

2. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: The pulse current sensor is connected to the grounding down conductor of the switch cabinet, and the ultrasonic sensor is attached to the insulating baffle of the switch cabinet to generate a synchronous current pulse sequence and ultrasonic waveform, including: The high-frequency current signal in the grounding down conductor is collected by a pulse current sensor and a standardized current pulse sequence is generated after filtering. The ultrasonic sensor receives the sound wave signal inside the switch cabinet and generates an ultrasonic waveform after noise suppression processing; A time synchronization device is used to perform clock alignment on the current pulse sequence and the ultrasonic waveform.

3. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: The calculating, based on the current pulse sequence and the ultrasonic waveform, the signal delay difference between the ground down conductor and the sensor at the insulating barrier comprises: Extracting the first pulse peak moment of the current pulse sequence as a reference moment; Identifying a starting fluctuation point in the ultrasonic waveform corresponding to the reference time; The time difference between the reference time and the starting fluctuation point is calculated as the initial delay difference.

4. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: The method of establishing a multi-path acoustic wave propagation database including a direct path and an expected reflection path based on the three-dimensional structure of the metal cavity of the switch cabinet and the sensor layout coordinates includes: Measure the inner wall dimensions, corner angles, and insulation baffle positions of the switchgear metal cavity to build a three-dimensional structural model; Marking the layout coordinates of the pulse current sensor and the ultrasonic sensor in the three-dimensional structural model to determine the straight-line distance between the sensors; Simulating a direct path of an acoustic wave from a discharge point to the ultrasonic sensor, and recording the path length and propagation medium properties; Simulate the expected reflection path of the sound wave after it is reflected by the metal cavity wall and the insulating baffle, and record the number of reflections, the coordinates of the reflection point, and the path length; The parameters of the direct path and the expected reflection path are summarized to establish a multi-path database for acoustic wave propagation.

5. The method for locating and detecting partial discharge of electrical equipment according to claim 4, wherein: The direct path of the simulated sound wave from the discharge point to the ultrasonic sensor, and recording of the path length and propagation medium properties, include: Setting a virtual discharge point set in the three-dimensional structural model, wherein the virtual discharge point set covers a potential discharge area of ​​the metal cavity of the switch cabinet; For each virtual discharge point, a straight line propagation trajectory from the virtual discharge point to the ultrasonic sensor is generated as a direct path, and the path length of the straight line propagation trajectory and the sound velocity characteristics of the propagation medium are recorded.

6. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: The matching of the first arrival wave feature identified in the ultrasonic waveform with the acoustic wave propagation multi-path database to determine the actual acoustic wave propagation path type and the corresponding equivalent propagation distance includes: Extracting the peak amplitude, rising edge slope and waveform duration of the first arriving wave from the ultrasonic waveform as the arriving wave characteristic parameters; Calling the theoretical waveform characteristic parameters corresponding to each path in the acoustic wave propagation multi-path database; Calculate the similarity between the characteristic parameters of the first arrival wave and the characteristic parameters of each theoretical waveform; Based on the path corresponding to the theoretical waveform with the highest similarity, the actual sound wave propagation path type and equivalent propagation distance are determined.

7. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: The iteratively determining the optimal spatial coordinates of the discharge point by using an optimization algorithm in combination with the signal delay difference, the sensor layout coordinates, and the equivalent propagation distance includes: Using the layout coordinates of the pulse current sensor and the ultrasonic sensor as the boundary points, a cubic space containing the line connecting the two and the surrounding area is defined as the initial spatial coordinate range of the discharge point; Determining a distance constraint condition between the discharge point and the two sensors based on the signal delay difference and the equivalent propagation distance; Based on the distance constraint, constructing an objective function with the spatial coordinates of the discharge point as variables, wherein the output value of the objective function is the sum of the deviations of the distance constraint; Iteratively calculating the objective function using a particle swarm optimization algorithm, and updating the candidate coordinates of the discharge point in each iteration; When the calculation error of the candidate coordinates is less than a preset threshold, the coordinates are output as the optimal spatial coordinates.

8. The method for locating and detecting partial discharge of electrical equipment according to claim 7, wherein: The particle swarm optimization algorithm is used to iteratively calculate the objective function, and each iteration updates the candidate coordinates of the discharge point, including: Initialize the particle positions of the particle swarm, where each particle corresponds to a candidate coordinate of the discharge point, and the particle positions are distributed within the initial spatial coordinate range; Taking the output value of the objective function as fitness, updating the speed and position of the particles through the particle swarm optimization algorithm, and iteratively optimizing the candidate coordinates; Calculate the output value of the objective function corresponding to the candidate coordinates after each iteration, and use the output value as the calculation error.

9. The method for locating and detecting partial discharge of electrical equipment according to claim 7, wherein: The signal delay difference is used to correct the equivalent propagation distance, specifically including: Calculating a time correction value for sound wave propagation based on the signal delay difference; Based on the sound velocity of the air medium in the switch cabinet, multiplying the time correction value by the sound velocity to obtain a distance correction value; The distance correction value is subtracted from the equivalent propagation distance to obtain a corrected actual propagation distance, wherein the corrected actual propagation distance is used as an input parameter for iterative solution of the optimization algorithm.

10. The method for locating and detecting partial discharge of electrical equipment according to claim 1, wherein: Mapping the optimal spatial coordinates to a switchgear structural topology diagram to generate a partial discharge location report for electrical equipment includes: Converting the optimal spatial coordinates into two-dimensional coordinates in a switch cabinet structure topology diagram; Mark the internal components of the switch cabinet corresponding to the coordinates, including busbars, insulators, and circuit breakers; Count the occurrence frequency and discharge intensity of discharge points within the preset time; A positioning report including positioning position, discharge intensity and potential fault type is generated in combination with component attributes of the internal components of the switch cabinet.

Images