Method for determining insulation state of power distribution network equipment

Abstract

The embodiment of the invention discloses a method for determining the insulation state of power distribution network equipment. The method comprises the following steps: drawing an insulation state judgment ball of the power distribution network equipment in a three-dimensional rectangular coordinate system in an interface of computer equipment; wherein a normal area, an attention area and a dangerous area are divided and displayed in the insulation state judgment ball; acquiring three target discharge characteristic parameters of a target discharge source of the target power distribution network equipment in a preset time period; according to the three target discharge characteristic parameters, determining target three-dimensional coordinates of a target discharge source in the three-dimensional rectangular coordinate system; and drawing a point corresponding to the target three-dimensional coordinate in an insulation state judgment ball, and determining the insulation state of the target power distribution network equipment according to an area in which the point corresponding to the target three-dimensional coordinate falls. Therefore, comprehensive consideration of discharge characteristic parameters of multiple dimensions is realized, limitation of traditional single parameter evaluation is overcome, and full utilization and analysis of partial discharge data are realized.

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, and in particular to a method for determining the insulation status of power distribution network equipment. Background Technology

[0002] Currently, with the widespread application of sensor technology in power distribution networks, massive amounts of partial discharge data can be collected online or offline. However, due to the lack of intelligent diagnostic methods for the insulation status of power distribution network equipment based on partial discharge data, this partial discharge data cannot be fully explored and utilized. Specifically, most existing analysis methods remain at the stage of simple threshold alarms and manual spectrum interpretation, making it difficult to provide quantitative basis for assessing insulation degradation.

[0003] Therefore, there is a need for an intelligent diagnostic method for the insulation status of distribution network equipment that can make full use of partial discharge data, so as to make full use of partial discharge data and improve the diagnostic effectiveness of the insulation status of distribution network equipment. Summary of the Invention

[0004] The main objective of this invention is to provide a method, apparatus, computer equipment, and storage medium for determining the insulation status of power distribution network equipment. This invention can solve the problems in the prior art of lacking an intelligent diagnostic method for the insulation status of power distribution network equipment based on partial discharge data, and the low diagnostic effectiveness of the insulation status of power distribution network equipment.

[0005] To achieve the above objectives, the first aspect of the present invention provides a method for determining the insulation status of power distribution network equipment, applied to computer equipment, the method comprising: A three-dimensional rectangular coordinate system is constructed, and the three discharge characteristic parameters of the power supply of the power distribution network equipment are mapped to the three-dimensional rectangular coordinate system as x, y, and z axes, respectively, in order to construct a spherical equation based on the discharge characteristic vector; Based on the sphere equation, an insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system is drawn on the interface of the computer device; wherein, the insulation status judgment sphere is divided into normal area, attention area and danger area. Obtain three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period; Based on the three target discharge characteristic parameters, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system. The point corresponding to the three-dimensional coordinates of the target is plotted on the insulation state judgment sphere. The insulation state of the target power distribution equipment is determined based on the area where the point corresponding to the three-dimensional coordinates of the target falls.

[0006] In some embodiments, the above-mentioned drawing of the insulation state judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system in the interface of the computer device according to the sphere equation includes: Obtain the modulus range of each insulation state level obtained by classifying the insulation state of the power distribution network equipment; wherein, the multiple insulation state levels include normal level, warning level and dangerous level; the modulus range is the range of the modulus of the three-dimensional coordinates determined by the three discharge characteristic parameters of the first discharge source, and the insulation state level of the first discharge source is the insulation state level corresponding to the modulus range. For each insulation state level, the maximum value in the modulus division range of the insulation state level is used as the radius of the sphere equation to obtain the target sphere equation corresponding to the insulation state level. Based on the target sphere equation corresponding to each insulation state level, the target sphere corresponding to each insulation state level in the three-dimensional rectangular coordinate system is drawn in the interface of the computer device. The sphere range corresponding to the normal level is displayed as the normal area, the sphere range corresponding to the attention level minus the sphere range corresponding to the normal level is displayed as the attention area, and the sphere range corresponding to the danger level minus the sphere range corresponding to the attention level is displayed as the danger area, thus obtaining the insulation state judgment sphere.

[0007] In some embodiments, after plotting the point corresponding to the target three-dimensional coordinates on the insulation state judgment sphere, the method includes: connecting the point corresponding to the target three-dimensional coordinates with historical adjacent points to obtain the connection curve of the target discharge source; the historical adjacent points are points that are temporally adjacent to the point corresponding to the target three-dimensional coordinates among the historical points corresponding to the target discharge source, and two temporally adjacent points among the historical points corresponding to the target three-dimensional coordinates are connected; when it is determined, based on the connection curve of the target discharge source, that the point corresponding to the target discharge source is gradually moving towards the attention area or danger area in the insulation state judgment sphere, the user is reminded to take intervention measures for the power distribution network equipment.

[0008] In some embodiments, obtaining the modulus range of each insulation state level obtained by classifying the insulation state of the distribution network equipment includes: obtaining the level scoring vector corresponding to each insulation state level; obtaining the value range of the fuzzy evaluation vector and the value range of the feature parameter weight vector corresponding to the three discharge feature parameters of the first discharge source; the fuzzy evaluation vector is used to characterize the degree of membership of the quantized value corresponding to the discharge feature parameter to the normal level, the attention level, and the danger level, respectively; multiplying the level scoring vector and the value range of the fuzzy evaluation vector corresponding to each insulation state level to obtain a first value range of the standardized value corresponding to each insulation state level; multiplying the first value range and the value range of the feature parameter weight vector to obtain a second value range of the coordinate values ​​of each insulation state level in three-dimensional coordinates; calculating the modulus using the second value range of each insulation state level in three-dimensional coordinates to obtain the modulus value range composed of the modulus of the coordinate values ​​of each insulation state level in three-dimensional coordinates; and using the modulus value range corresponding to each insulation state level as the modulus range of each insulation state level.

[0009] In some embodiments, the modulus of the three-dimensional coordinates is calculated according to the following formula: ; in, Represents the coordinate values ​​of three-dimensional coordinates. The modulus of the three-dimensional coordinate system.

[0010] In some embodiments, determining the target three-dimensional coordinates of the target discharge source in the three-dimensional Cartesian coordinate system based on the three target discharge feature parameters includes: quantizing the three target discharge feature parameters to obtain target quantized values ​​corresponding to the three target discharge feature parameters; for each target discharge feature parameter, determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantized value corresponding to the target discharge feature parameter; for each target discharge feature parameter, determining the target coordinate value corresponding to the target discharge feature parameter by multiplying the target fuzzy evaluation vector corresponding to the target discharge feature parameter, a preset level scoring vector, and a preset feature parameter weight; and determining the target three-dimensional coordinates of the target discharge source in the three-dimensional Cartesian coordinate system based on each target coordinate value.

[0011] In some embodiments, the above method further includes: Obtain the judgment matrix corresponding to the three discharge feature parameters; wherein, the judgment matrix includes the importance ratio between each pair of discharge feature parameters among the three discharge feature parameters; The judgment matrix is ​​column-wise normalized using a first normalization formula to obtain a normalized matrix; wherein, the first normalization formula is: bij =a ij / Σa ij ; Among them, b ij Let a represent the element in the i-th row and j-th column of the normalized matrix. ij Σa represents the element in the i-th row and j-th column of the judgment matrix. ij This represents the sum of the elements in the j-th column of the judgment matrix; The sum of each row of the normalized matrix is ​​calculated to obtain the first summation value, the second summation value, and the third summation value. The second normalization formula is used to normalize the first summation value, the second summation value, and the third summation value respectively, to obtain the preset feature parameter weights corresponding to the three discharge feature parameters; the second normalization formula is as follows: v i = w i / Σ w i ; in, v i This represents the weight of the i-th preset feature parameter. w i Σ represents the summation value of the i-th term, where i takes values ​​from 1 to 3. w i This represents the sum of the first summation value, the second summation value, and the third summation value.

[0012] In some embodiments, determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter includes: determining a first membership degree, a second membership degree, and a third membership degree corresponding to the target quantization value from a preset membership degree rule based on the target quantization value corresponding to the target discharge feature parameter; wherein, the first membership degree is the membership degree of the target quantization value belonging to the normal level, the second membership degree is the membership degree of the target quantization value belonging to the attention level, and the third membership degree is the membership degree of the target quantization value belonging to the danger level, and the membership degree rule includes the correspondence between different quantization values ​​of different discharge feature parameters and membership degrees; and determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the first membership degree, the second membership degree, and the third membership degree.

[0013] In some embodiments, determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter includes: Obtain the set of historical target discharge feature parameters corresponding to the target discharge feature parameters; wherein, the set of historical target discharge feature parameters includes multiple historical target discharge feature parameters; The K-Means artificial intelligence clustering algorithm was used to process the set of historical target discharge feature parameters to obtain the center values ​​of three clusters; Based on the center values ​​of the three clusters, triangular membership functions are constructed for the target discharge characteristic parameters; the triangular membership functions include normal membership functions, attention membership functions, and danger membership functions; Substituting the target quantized value into the normal membership function yields the first membership degree; substituting the target quantized value into the attention membership function yields the second membership degree; and substituting the target quantized value into the danger membership function yields the third membership degree. Based on the first membership degree, the second membership degree, and the third membership degree, determine the fuzzy evaluation vector corresponding to the target discharge characteristic parameter; The expression for the trigonometric membership function is as follows: ; ; Where x is the target discharge characteristic parameter, μ k μ(x) is the membership function, k is a constant, taking values ​​of 1, 2, and 3 respectively. When k is 1, μ1(x) is the normal membership function; when k is 2, μ2(x) is the attention membership function; when k is 3, μ3(x) is the attention membership function; b1 is the minimum value among the center values ​​of the three clusters; b3 is the maximum value among the center values ​​of the three clusters; b2 is the remaining center value after removing the minimum and maximum values ​​from the center values ​​of the three clusters; and D is the set of historical target discharge characteristic parameters.

[0014] To achieve the above objectives, a second aspect of the present invention provides an apparatus for determining the insulation status of power distribution network equipment, comprising a computer device, the apparatus comprising: The module is used to construct a three-dimensional rectangular coordinate system, which maps the three discharge characteristic parameters of the power supply of the power distribution network equipment as x, y, and z axes to the three-dimensional rectangular coordinate system, so as to construct a spherical equation based on the discharge characteristic vector. The drawing module is used to draw the insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system in the interface of the computer device according to the sphere equation; wherein the insulation status judgment sphere is divided into normal area, attention area and danger area. The parameter acquisition module is used to acquire three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period; The coordinate determination module is used to determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system based on the three target discharge characteristic parameters. The state determination module is used to draw the point corresponding to the three-dimensional coordinates of the target on the insulation state judgment sphere, and determine the insulation state of the target power distribution equipment based on the area where the point corresponding to the three-dimensional coordinates of the target falls.

[0015] To achieve the above objectives, a third aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: A three-dimensional rectangular coordinate system is constructed, and the three discharge characteristic parameters of the power supply of the power distribution network equipment are mapped to the three-dimensional rectangular coordinate system as x, y, and z axes, respectively, in order to construct a spherical equation based on the discharge characteristic vector; Based on the sphere equation, an insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system is drawn on the interface of the computer device; wherein, the insulation status judgment sphere is divided into normal area, attention area and danger area. Obtain three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period; Based on the three target discharge characteristic parameters, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system. The point corresponding to the three-dimensional coordinates of the target is plotted on the insulation state judgment sphere. The insulation state of the target power distribution equipment is determined based on the area where the point corresponding to the three-dimensional coordinates of the target falls.

[0016] To achieve the above objectives, a fourth aspect of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor causes the processor to perform the following steps: A three-dimensional rectangular coordinate system is constructed, and the three discharge characteristic parameters of the power supply of the power distribution network equipment are mapped to the three-dimensional rectangular coordinate system as x, y, and z axes, respectively, in order to construct a spherical equation based on the discharge characteristic vector; Based on the sphere equation, an insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system is drawn on the interface of the computer device; wherein, the insulation status judgment sphere is divided into normal area, attention area and danger area. Obtain three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period; Based on the three target discharge characteristic parameters, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system. The point corresponding to the three-dimensional coordinates of the target is plotted on the insulation state judgment sphere. The insulation state of the target power distribution equipment is determined based on the area where the point corresponding to the three-dimensional coordinates of the target falls.

[0017] The embodiments of the present invention have the following beneficial effects: This invention provides a method, apparatus, computer equipment, and storage medium for determining the insulation status of distribution network equipment. By acquiring multiple target discharge characteristic parameters and determining the target three-dimensional coordinates based on the multiple target discharge characteristic parameters, the insulation status of the distribution network equipment is determined. This invention can comprehensively consider discharge characteristic parameters from multiple dimensions, overcome the limitations of traditional single-parameter evaluation, significantly improve the comprehensiveness and accuracy of status evaluation, and realize the full utilization and analysis of partial discharge data. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] in: Figure 1 This is a flowchart illustrating the method for determining the insulation status of power distribution network equipment in an embodiment of the present invention. Figure 2 This is a schematic diagram of the insulation status judgment ball of the power distribution network equipment in an embodiment of the present invention; Figure 3 The trigonometric membership function for the single average discharge amplitude in the embodiments of the present invention; Figure 4 This is the trigonometric membership function of the discharge frequency in the embodiments of the present invention; Figure 5 A schematic diagram of an insulation state judgment sphere based on coordinate points is shown in an embodiment of the present invention; Figure 6 This is a flowchart illustrating the practical operation of the comprehensive evaluation method for the insulation status of power distribution network equipment in this embodiment of the invention. Figure 7 This is a schematic diagram of the voltage across the impedance versus time in an embodiment of the present invention; Figure 8(a) shows the PRPS spectrum before denoising in an embodiment of the present invention; Figure 8(b) shows the denoised PRPS spectrum in an embodiment of the present invention; Figure 9 This is a visualization of the continuous multi-day monitoring results of the equipment insulation status in an embodiment of the present invention; Figure 10 This is a structural block diagram of the device for determining the insulation status of power distribution network equipment in an embodiment of the present invention; Figure 11 This is a structural block diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In one embodiment, a method for determining the insulation status of power distribution network equipment is provided, such as... Figure 1 As shown, Figure 1 This is a flowchart illustrating a method for determining the insulation state of power distribution network equipment according to an embodiment of the present invention. This method can be applied to computer equipment, which can be a terminal or a server. The terminal can be a computer, mobile phone, tablet, etc. This embodiment illustrates the application to a terminal. The method for determining the insulation state of power distribution network equipment specifically includes the following steps S101-S105: Step S101: Construct a three-dimensional rectangular coordinate system. Map the three discharge characteristic parameters of the power supply of the power distribution network equipment as x, y, and z axes to the three-dimensional rectangular coordinate system to construct a spherical equation based on the discharge characteristic vector.

[0022] A discharge source can refer to a physically distinguishable specific insulation defect point or area that causes continuous or repetitive partial discharge inside or on the surface of the insulation system of electrical equipment (distribution network equipment) due to the concentration of electric field. Multiple discharge sources can exist on a single distribution network device.

[0023] The aforementioned discharge characteristic parameters are indicators used to diagnose the severity of insulation defects caused by discharge sources. These discharge characteristic parameters are key indicators that quantitatively describe the discharge mode and severity, extracted from raw partial discharge data through a series of algorithms including signal processing, statistical analysis, and feature extraction.

[0024] In this embodiment, the three discharge characteristic parameters of the power supply of the distribution network equipment are respectively used as variables of the x, y, and z axes of a three-dimensional rectangular coordinate system. In one implementation, the three discharge characteristic parameters may include discharge type, discharge frequency, and average discharge amount per discharge, wherein the discharge type, discharge frequency, and average discharge amount per discharge are respectively variables of the x, y, and z axes.

[0025] The above-mentioned spherical equation based on the discharge characteristic vector can be represented by a circle centered at the origin of a three-dimensional rectangular coordinate system; that is, the spherical equation can be... x 2 + y 2 + z2 = r 2 .

[0026] Step S102: Based on the sphere equation, draw the insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system on the computer device interface.

[0027] The insulation condition assessment sphere is divided into normal area, warning area, and danger area.

[0028] In some embodiments, step S102 may specifically include the following steps: Obtain the modulus range of each insulation state level among multiple insulation state levels for classifying the insulation state of distribution network equipment.

[0029] For each insulation condition level, the maximum value in the range of the modulus division of the insulation condition level is used as the radius of the sphere equation to obtain the target sphere equation corresponding to the insulation condition level.

[0030] Based on the target sphere equation corresponding to each insulation state level, the target sphere corresponding to each insulation state level is drawn in the computer device interface in a three-dimensional rectangular coordinate system.

[0031] The sphere range corresponding to the normal level is displayed as the normal area, the sphere range corresponding to the attention level (excluding the sphere range corresponding to the normal level) is displayed as the attention area, and the sphere range corresponding to the danger level (excluding the sphere range corresponding to the attention level) is displayed as the danger area, thus obtaining the insulation status judgment sphere.

[0032] The aforementioned insulation condition levels may include normal, caution, and hazardous levels.

[0033] For the modulus length division range of each insulation state level, the modulus length division range is the range of the three-dimensional coordinates where the modulus is located, determined by the three discharge characteristic parameters of the first discharge source. The insulation state level of the first discharge source is the insulation state level corresponding to the modulus length division range. For example, when the range of the modulus of the three-dimensional coordinates corresponding to the discharge source is [0, ... When the corresponding insulation condition level is normal, the modulus range corresponding to insulation condition level I (normal level) is [0, When the range of the magnitude of the three-dimensional coordinates corresponding to the discharge source is ( , The corresponding insulation condition level is the caution level. Therefore, the modulus range corresponding to insulation condition level II (caution level) is ( , When the range of the magnitude of the three-dimensional coordinates corresponding to the discharge source is ( , The corresponding insulation condition level is hazardous. Therefore, the modulus range corresponding to insulation condition level III (hazardous level) is ( , ].

[0034] For each insulation condition level, the maximum value within the modulus range of the insulation condition level is used as the radius of the sphere equation to obtain the target sphere equation corresponding to the insulation condition level. For example: Insulation condition level I (normal level) corresponds to a modulus range of [0, ...]. The modulus range corresponding to insulation condition class II (caution level) is ( , The modulus range corresponding to insulation condition class III (hazardous class) is ( , At this point, the maximum value of the modulus division range corresponding to the normal level is... Note that the maximum value of the modulus range corresponding to the level is The maximum value of the modulus range corresponding to the danger level is For the normal level, r = The equation of the target sphere corresponding to the normal level is: For the attention level, r = Note that the equation of the target sphere corresponding to the level is: For the danger level, r = The equation of the target sphere corresponding to the danger level is: .

[0035] In this embodiment, by setting the maximum value of the pre-determined modulus division range used for diagnosing insulation condition levels as the radius of the judgment sphere, the boundary of the three-dimensional visualized sphere is made completely consistent with the diagnostic criteria for insulation condition. This design ensures that the judgment sphere is not a simple graphic detached from actual data, but a spatial reproduction of diagnostic logic, thereby significantly improving the effectiveness and reliability of the judgment sphere generation, and further enhancing the effectiveness of the results intuitively judged by maintenance personnel based on the sphere area.

[0036] In some embodiments, the determination of the above-mentioned three-dimensional coordinates can be achieved through the following steps: The three discharge characteristic parameters are quantized to obtain their corresponding quantized values.

[0037] For each discharge characteristic parameter, the fuzzy evaluation vector corresponding to the discharge characteristic parameter is determined based on the quantization value corresponding to the discharge characteristic parameter. The product of the fuzzy evaluation vector corresponding to the discharge characteristic parameter and the preset level scoring vector is determined as the standardized quantization value corresponding to the discharge characteristic parameter. The product of the standardized quantization value of the discharge characteristic parameter and the preset characteristic parameter weight corresponding to the discharge characteristic parameter is determined as the coordinate value corresponding to the discharge characteristic parameter.

[0038] Among them, the fuzzy evaluation vector is used to characterize the degree of membership of the quantized value corresponding to the discharge characteristic parameter to the normal level, the attention level, and the danger level, respectively.

[0039] Based on the coordinate values, determine the three-dimensional coordinates of the power source in the three-dimensional rectangular coordinate system.

[0040] The above three-dimensional coordinates P( ) modulus The calculation formula is: .

[0041] Based on the rules for determining three-dimensional coordinates and modulus, in some embodiments, obtaining the modulus division range for each insulation state level of the distribution network equipment insulation state classification may include the following steps: Obtain the grade score vector corresponding to each insulation condition level. The grade score vector includes the grade score corresponding to each insulation condition level.

[0042] The range of values ​​for the fuzzy evaluation vectors corresponding to the three discharge characteristic parameters of the first discharge source and the range of values ​​for the characteristic parameter weight vectors are obtained. The fuzzy evaluation vectors are used to characterize the degree of membership of the quantized values ​​corresponding to the discharge characteristic parameters to the normal level, the attention level and the danger level, respectively.

[0043] Multiply the value ranges of the grade scoring vector and the fuzzy evaluation vector corresponding to each insulation state level to obtain the first value range of the standardized quantitative value corresponding to each insulation state level.

[0044] Multiply the first value range with the value range of the feature parameter weight vector to obtain the second value range of the coordinate values ​​of each insulation state level in three-dimensional coordinates; The modulus is calculated by using the second value range of each insulation condition level in three-dimensional coordinates. The modulus value range formed by the modulus of the coordinate values ​​of each insulation condition level in three-dimensional coordinates is obtained, and the modulus value range corresponding to each insulation condition level is used as the modulus division range of each insulation condition level.

[0045] For example, the grade scoring vector corresponding to each insulation condition level is [1,2,3]. T 1, 2, and 3 represent the rating levels for normal, attention, and danger levels, respectively, and are fuzzy evaluation vectors.R i =[ r i1 , r i2 , r i3 The sum of all elements in the vector is equal to 1, so the range of values ​​for each element in the fuzzy evaluation vector is [0,1]. The most extreme value of the fuzzy evaluation vector corresponding to the three discharge characteristic parameters of the first discharge source in the normal stage is [0,1]. 1 , 0 , 0 The most extreme value of the fuzzy evaluation vector corresponding to the three discharge characteristic parameters of the first discharge source of the attention stage is []. 0 , 1 , 0 The most extreme value of the fuzzy evaluation vector corresponding to the three discharge characteristic parameters of the first discharge source of the hazardous level is []. 0 , 0 , 1 The rating vector is [1,2,3]. T ， For the normal level, the maximum value of Si is 1, and the minimum value is 0, meaning the first value range is [0, 1]. Also, the feature parameter weight vector... V The sum of all elements equals 1, and the values ​​of each element in the feature parameter weight vector range from [0,1]. x 0 =s x × v 1、 y 0= s y × v 2. Z 0= s z × v 3， The value range is within [0, 1], meaning the second value range is [0, 1]. The largest coordinate is (1, 1, 1), and the smallest coordinate is (0, 0, 0). Therefore, the modulus value range is [0, 1]. Within the range of [0, 1], for the attention level, the maximum value of Si is 2. Since the value range of each element of the feature parameter weight vector is [0, 1], the maximum value of Si is 2, and the minimum value is 0, meaning the first value range is [0, 2]. The feature parameter weight vector... V The sum of all elements equals 1, and the values ​​of each element in the feature parameter weight vector range from [0,1]. x 0 =s x × v 1、y 0= s y × v 2. Z 0= s z × v 3， The value range is [0, 2], and the second value range corresponding to the normal level is [0, 1]. Therefore, the second value range corresponding to the level is (1, 2]. The largest coordinate is (2, 2, 2), and the smallest coordinate is (1, 1, 1). Thus, the modulus value range is ( ). , For the danger level, the maximum value of Si is 3, meaning the first value range is [0, 3]. Also, the characteristic parameter weight vector... V The sum of all elements equals 1, and the values ​​of each element in the feature parameter weight vector range from [0,1]. x 0 =s x × v 1、 y 0= s y × v 2. Z 0= s z × v 3， The value range is within [0, 3]. Note that the second value range corresponding to the first level is (1, 2]. Therefore, the second value range corresponding to the second level is (2, 3]. The largest coordinate is (3, 3, 3), and the smallest coordinate is (2, 2, 2). Thus, the modulus range is (2, 3). 3 ].

[0046] After generating the insulation status judgment ball, the insulation status of the power distribution network equipment can be visually judged using the insulation status judgment ball.

[0047] The following steps S103-S105 describe the application of the insulation state judgment ball.

[0048] Step S103: Obtain three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period.

[0049] The aforementioned target distribution network equipment refers to the distribution network equipment whose insulation status needs to be determined, and the aforementioned target discharge source refers to the discharge source of the target distribution network equipment. It can refer to each discharge source of the target distribution network equipment. The determination of the insulation status is explained by taking the target discharge source as the description object.

[0050] The above-mentioned preset time period can be set according to needs. For example, the preset time period is one day in length, and the preset time period can be from 00:00 on November 1, 2025 to 00:00 on November 2, 2025.

[0051] The aforementioned target discharge characteristic parameters are the specific values ​​or types of the aforementioned discharge characteristic parameters.

[0052] In some embodiments, to improve the accuracy of model evaluation, this invention introduces a specialized discharge data denoising method before extracting discharge feature parameters. Step 103 above may include the following steps: Obtain the first partial discharge dataset monitored on the distribution network equipment within a preset time period.

[0053] The partial discharge data of at least one of the second and third discharge sources is removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal.

[0054] Three target discharge characteristic parameters of the target discharge source are extracted from the second partial discharge dataset.

[0055] The aforementioned first partial discharge dataset includes partial discharge data (hereinafter referred to as partial discharge data) from at least one discharge source. Partial discharge data refers to information collected from partial discharge phenomena through sensors and monitoring systems, which can be used to characterize discharge characteristics, identify defect types, and assess insulation status.

[0056] The number of discharge events occurring in the second discharge source within the preset time period shall not exceed the preset number, and the discharge frequency of the third discharge source shall exceed the preset frequency, but the single discharge amount shall be lower than the preset discharge amount.

[0057] The preset number of times can be set as required, for example, it can be set to 5 times, but it is not limited to 5 times.

[0058] The preset frequency can be set as required.

[0059] The preset discharge amount can be set as required, for example, it can be set to 5pC, but it is not limited to setting it to 5pC.

[0060] The above-mentioned partial discharge data of at least one of the second and third discharge sources are removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal. That is, the partial discharge data of the second discharge source is removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal, or the partial discharge data of the third discharge source is removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal, or the partial discharge data of both the second and third discharge sources is removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal.

[0061] For example, firstly, discharge sources that occur less than 5 times are eliminated to rule out the impact of occasional interference pulses on the evaluation results; secondly, discharge sources that occur frequently but whose single discharge quantity is always less than 5pC are eliminated to prevent continuous low-energy interference from misleading the equipment status judgment.

[0062] The target discharge source mentioned above is any discharge source included in the second partial discharge dataset.

[0063] In this embodiment, the above-mentioned denoising steps can effectively filter out interference components and retain real and valid partial discharge data, laying a solid foundation for subsequent feature extraction and evaluation.

[0064] Step S104: Determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system based on the three target discharge characteristic parameters.

[0065] In some embodiments, step S104 above may include the following steps: The three target discharge characteristic parameters are quantized to obtain the target quantized values ​​corresponding to the three target discharge characteristic parameters.

[0066] For each target discharge characteristic parameter, the target fuzzy evaluation vector corresponding to the target discharge characteristic parameter is determined based on the target quantization value corresponding to the target discharge characteristic parameter.

[0067] For each target discharge characteristic parameter, the product of the target fuzzy evaluation vector corresponding to the target discharge characteristic parameter, the preset level score vector, and the preset characteristic parameter weight is determined as the target coordinate value corresponding to the target discharge characteristic parameter. Based on the target coordinate values, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system.

[0068] Among them, the target fuzzy evaluation vector is used to characterize the degree of membership of the target quantization value corresponding to the target discharge characteristic parameter to the normal level, attention level and danger level, respectively.

[0069] The aforementioned preset feature parameter weights can be set as needed. Different preset feature parameter weights can be set for different discharge feature parameters. For example, the three discharge feature parameters can include discharge type, discharge frequency, and average discharge amount per discharge, with the preset feature parameter weight corresponding to the discharge type. It can be set to 0.668, which is the preset feature parameter weight corresponding to the discharge frequency. It can be set to 0.243, which is the preset feature parameter weight corresponding to the average discharge quantity per cycle. It can be set to 0.088.

[0070] The aforementioned preset feature parameter weights can also be obtained through calculation.

[0071] The above method of determining the target three-dimensional coordinates of the target discharge source in a three-dimensional rectangular coordinate system based on the coordinate values ​​corresponding to the discharge characteristic parameters of each target can be used to compose the target three-dimensional coordinates.

[0072] For example, the three discharge characteristic parameters may include discharge type, discharge frequency, and average discharge quantity per discharge. The coordinates corresponding to the discharge type are the coordinates on the x-axis of the three-dimensional coordinate system, the coordinates corresponding to the discharge frequency are the coordinates on the y-axis of the three-dimensional coordinate system, and the coordinates corresponding to the average discharge quantity per discharge are the coordinates on the z-axis of the three-dimensional coordinate system. The fuzzy evaluation vector R corresponding to the target discharge characteristic parameters... i =[r i1 r i2 r i3 The rating vector is [1,2,3]. T The fuzzy evaluation vector R corresponding to the target discharge characteristic parameters i =[r i1 r i2 r i3 [1,2,3] and the rating vector. T The products are multiplied and then weighted to obtain a unique standardized quantized value s. i ,Right now: ; In the formula, r ij The target discharge characteristic parameter x i For evaluation level L j The degree of membership is determined by the values ​​of i and j, which range from 1 to 3, and both i and j are positive integers. L1 is the normal level, L2 is the attention level, and L3 is the danger level.

[0073] The coordinate value on the x-axis corresponding to the target discharge type for × The y-coordinate value corresponding to the target discharge frequency for × The z-axis coordinate value corresponding to the target average discharge quantity per cycle for × , , , The above are respectively , , The analytic hierarchy process (AHP) confirmed the discharge characteristic parameter weight vector as V = [0.668, 0.243, 0.088]. After standardization and weighting, the discharge characteristic parameters were transformed into [x0, y0, z0] = [...]. × , × , × ], which correspond to the actual values ​​of the x, y, and z axes, respectively.

[0074] In this embodiment, by standardizing discharge characteristic parameters of different dimensions and ranges, the influence of parameter magnitude differences on the evaluation results is eliminated, laying a technical foundation for the effective fusion of multiple discharge characteristic parameters. Furthermore, by multiplying the preset characteristic parameter weights with the standardized quantization values, the physical meaning of the original discharge characteristic parameters is preserved, while also reflecting the relative importance of different discharge characteristic parameters in the condition assessment. This allows the constructed target three-dimensional coordinates to more accurately reflect the true insulation state of the distribution network equipment.

[0075] In some embodiments, the quantization range for quantizing the three discharge characteristic parameters of the discharge power source can be set to 0~1, with higher values ​​indicating more severe insulation defects.

[0076] The above-mentioned quantization processing of the three target discharge characteristic parameters to obtain the target quantized values ​​corresponding to the three target discharge characteristic parameters can be performed on the non-numerical target discharge characteristic parameters to obtain the target value quantized values ​​of the non-numerical target discharge characteristic parameters, and the numerical target discharge characteristic parameters are determined as the target value quantized values ​​corresponding to the numerical target discharge characteristic parameters. Therefore, the target quantized values ​​corresponding to the three target discharge characteristic parameters include the target value quantized values ​​of the numerical target discharge characteristic parameters and the numerical target discharge characteristic parameters.

[0077] For example, if the three target discharge characteristic parameters include x1 = tip discharge (target discharge type), x2 = 45 pC (target average discharge quantity per single charge), and x3 = 3 (target discharge frequency), and the discharge type is a non-numerical discharge characteristic parameter, while the average discharge quantity per single charge and the discharge frequency are numerical discharge characteristic parameters, then the tip discharge is quantized to obtain a target value of 0.1 for the tip discharge, and target values ​​of 45 and 3 for the average discharge quantity per single charge and the discharge frequency, respectively.

[0078] The above-described quantization process for the non-numerical target discharge characteristic parameters among the three target discharge characteristic parameters yields their target quantized values. This can be achieved by determining the corresponding target fault consequence value for each non-numerical target discharge characteristic parameter from a pre-defined quantization rule. The pre-defined quantization rule includes the correspondence between different categories of discharge characteristic parameters and fault consequence values, whereby the fault consequence value characterizes the severity of the potential consequences caused by the discharge characteristic parameter.

[0079] For example, the target discharge characteristic parameter is a tip discharge, where the discharge type is divided into tip discharge, suspended discharge, metal particle discharge, internal discharge, and surface discharge. The preset quantization rules corresponding to the discharge type can be tip discharge -0.1, suspended discharge -0.1, metal particle discharge -0.7, internal discharge -0.7, and surface discharge -0.9. That is, the fault consequence value corresponding to tip discharge is 0.1, the fault consequence value corresponding to suspended discharge is 0.1, the fault consequence value corresponding to metal particle discharge is 0.7, the fault consequence value corresponding to internal discharge is 0.7, and the fault consequence value corresponding to surface discharge is 0.9. When the target discharge type is tip discharge, the corresponding target fault consequence value is determined to be 0.1 from the corresponding preset quantization rules, that is, the target discharge characteristic parameter value after quantization of tip discharge is 0.1.

[0080] In some embodiments, determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter may include the following steps: Based on the target quantization value corresponding to the target discharge characteristic parameter, the first membership degree, second membership degree, and third membership degree corresponding to the target quantization value are determined from the preset membership degree rules.

[0081] Among them, the first membership degree is the membership degree of the target quantization value belonging to the normal level, the second membership degree is the membership degree of the target quantization value belonging to the attention level, and the third membership degree is the membership degree of the target quantization value belonging to the danger level. The membership degree rules include the correspondence between different quantization values ​​of different discharge characteristic parameters and membership degrees.

[0082] Based on the first membership degree, the second membership degree, and the third membership degree, the target fuzzy evaluation vector corresponding to the target discharge characteristic parameters is determined.

[0083] The above fuzzy evaluation vector can be R i =[r i1 r i2 r i3 ], r i1 r i2 r i3These are the first membership degree, the second membership degree, and the third membership degree, respectively.

[0084] For example, the membership rule includes the correspondence between different quantization values ​​of discharge types and membership degrees, as shown in the following 5×3 membership matrix, which corresponds to the membership degree distribution of the five discharge types at the three state levels.

[0085] ; When a point discharge occurs, its fuzzy evaluation vector R1 = [0.7, 0.2, 0.1], meaning the membership degree for the "normal" level is 0.7, indicating that point discharges are mostly in a normal state. The "caution" level is 0.2, and the "danger" level is 0.1, indicating that it is only classified as a dangerous state in very rare cases. When a suspended discharge occurs, its fuzzy evaluation vector is also R1 = [0.7, 0.2, 0.1], indicating that suspended discharges are also mainly in a normal state, with very few cases being dangerous. The fuzzy evaluation vector for metal particle discharge is R1 = [0.1, 0.2, 0.7], where the "normal" level is only 0.1, indicating that this type is mostly in a dangerous state. When an internal discharge occurs, its fuzzy evaluation vector is also R1 = [0.1, 0.2, 0.7], indicating that this type is mostly in a dangerous state. The fuzzy evaluation vector for surface discharge is R1=[0.0, 0.1, 0.9], meaning that the membership degree of the "normal" level is 0, and almost all of them belong to the dangerous level, indicating that this type of discharge has the greatest impact on the insulation safety of the equipment.

[0086] In this embodiment, the target discharge characteristic parameter can be the target discharge type. That is, when the target discharge characteristic parameter is the target discharge type, the target fuzzy evaluation vector of the target discharge type can be determined by the fuzzy evaluation vector determination method of this embodiment.

[0087] In another embodiment, the above-mentioned determination of the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter may further include the following steps: Obtain the set of historical target discharge characteristic parameters. This set, D, includes multiple historical target discharge characteristic parameters. For example, d i This represents the average discharge amount of the i-th historical target in a single instance.

[0088] Choosing K=3, the K-Means artificial intelligence clustering algorithm is used to process the set of historical target discharge feature parameters to obtain the center values ​​of three clusters. Specifically, existing techniques can be referenced, such as initially randomly selecting three center values, repeatedly iterating sample allocation and center updates until convergence, and finally obtaining the center values ​​of the three clusters.

[0089] Based on the center values ​​of the three clusters, a triangular membership function for the target discharge characteristic parameters is constructed.

[0090] The expression for the trigonometric membership function can be as follows: ; ; Where x is the target discharge characteristic parameter, μ k μ(x) is a membership function, and k is a constant, taking values ​​of 1, 2, and 3 respectively. When k is 1, μ1(x) is a normal membership function, used to determine the first membership degree corresponding to the target discharge characteristic parameter based on the target discharge characteristic parameter. When k is 2, μ2(x) is an attention membership function, used to determine the second membership degree corresponding to the target discharge characteristic parameter based on the target discharge characteristic parameter. When k is 3, μ3(x) is an attention membership function, used to determine the third membership degree corresponding to the target discharge characteristic parameter based on the target discharge characteristic parameter. b1 is the minimum value among the center values ​​of the three clusters, b3 is the maximum value among the center values ​​of the three clusters, b2 is the remaining center value after removing the minimum and maximum values ​​from the center values ​​of the three clusters, and D is the set of historical target discharge characteristic parameters.

[0091] Substituting the target discharge characteristic parameters into the normal membership function yields the first membership degree; substituting the target discharge characteristic parameters into the attention membership function yields the second membership degree; and substituting the target discharge characteristic parameters into the danger membership function yields the third membership degree.

[0092] Based on the first membership degree, the second membership degree, and the third membership degree, the fuzzy evaluation vector corresponding to the target discharge characteristic parameter is determined.

[0093] In this embodiment, the aforementioned target discharge characteristic parameters can be numerical discharge characteristic parameters, such as the target average discharge quantity per cycle or the target discharge frequency.

[0094] In some implementations, the method may further include the following steps: Obtain the judgment matrix corresponding to the three discharge characteristic parameters.

[0095] The judgment matrix is ​​normalized column-wise using the first normalization formula to obtain the normalized matrix.

[0096] Calculate the sum of each row of the normalized matrix to obtain the first, second, and third sum values.

[0097] The second normalization formula is used to normalize the first summation value, the second summation value, and the third summation value respectively, so as to obtain the preset characteristic parameter weights corresponding to the three discharge characteristic parameters.

[0098] The judgment matrix includes the importance ratios between each pair of the three discharge characteristic parameters. The three discharge characteristic parameters can include discharge type, average discharge quantity per instance, and discharge frequency. Specifically, the elements in the first row of the judgment matrix are the importance ratios of discharge type to discharge type, discharge type to average discharge quantity per instance, and discharge type to discharge frequency, respectively; the elements in the second row are the importance ratios of average discharge quantity per instance to discharge type, average discharge quantity per instance to average discharge quantity per instance, and average discharge quantity per instance to discharge frequency, respectively; and the elements in the third row are the importance ratios of discharge frequency to discharge type, discharge frequency to average discharge quantity per instance, and discharge frequency to discharge frequency, respectively.

[0099] For example, if the importance ratio of discharge type to average discharge quantity per instance is assigned to 3, the importance ratio of discharge type to discharge frequency is assigned to 7, and the importance ratio of average discharge quantity per instance to discharge frequency is assigned to 3, then the judgment matrix A can be as follows: .

[0100] The first normalization formula mentioned above can be expressed as follows: b ij =a ij / Σa ij ; Among them, b ij Let a represent the element in the i-th row and j-th column of the normalized matrix. ij Σa represents the element in the i-th row and j-th column of the matrix. ij This represents the sum of the elements in the j-th column of the judgment matrix.

[0101] for example, Normalize the judgment matrix A column-wise to obtain the normalized matrix B, i.e., b. ij =a ij / Σa ij Therefore, the normalized matrix B is: .

[0102] The above steps calculate the sum of each row of the normalized matrix, yielding the first, second, and third summations, as illustrated in the following example: Then, the first to third rows of the normalized matrix B are the first summation value w1=2.006, the second summation value w2=0.73, and the third summation value w3=0.265, respectively.

[0103] The second normalization formula can be as follows: v i = w i / Σ w i ; in, v i This represents the weight of the i-th preset feature parameter. w i Σ represents the summation value of the i-th term, where i takes values ​​from 1 to 3. w i This represents the sum of the first, second, and third sums.

[0104] The three discharge characteristic parameters may include discharge type, average discharge quantity per cycle, and discharge frequency. The preset characteristic parameter weight corresponding to discharge type can be the ratio of the first summation value to the target summation value. The preset characteristic parameter weight corresponding to average discharge quantity per cycle can be the ratio of the second summation value to the target summation value. The preset characteristic parameter weight corresponding to discharge frequency can be the ratio of the third summation value to the target summation value. The target summation value is the sum of the first summation value, the second summation value, and the third summation value.

[0105] The above normalization of the first, second, and third summations yields the preset feature parameter weights corresponding to the three discharge characteristic parameters. For example, the first summation w1 = 2.006, the second summation w2 = 0.73, and the third summation w3 = 0.265, followed by normalization... v i = w i / Σ w i The weights of each discharge characteristic parameter can then be obtained.

[0106] In some implementations, the above-mentioned second normalization formula is used to normalize the first summation value, the second summation value, and the third summation value respectively to obtain the preset feature parameter weights corresponding to the three discharge feature parameters, which may include the following steps: The second normalization formula is used to normalize the first summation value, the second summation value, and the third summation value respectively, so as to obtain the initial characteristic parameter weights corresponding to the three discharge characteristic parameters.

[0107] Based on the judgment matrix A and the transpose of the feature parameter weight vector V T Multiplying yields a new vector. AV T .

[0108] For the new vector AV TThe i-th component is divided by the i-th component in the transpose of the feature parameter weight vector. We obtain the i-th target ratio.

[0109] The mean of the n target ratios is determined as the largest eigenvalue. λ max .

[0110] The consistency index is determined based on the formula for calculating the consistency index, the largest eigenvalue, and n.

[0111] Based on n, obtain the corresponding target RI value from the analytic hierarchy process (AHP) standard table.

[0112] If the ratio of the consistency index to the target RI value is less than a preset threshold, the initial feature parameter weights corresponding to the three discharge feature parameters are determined as the target feature parameter weights corresponding to the three discharge feature parameters.

[0113] If the ratio of the consistency index to the target RI value is greater than or equal to a preset threshold, adjust the judgment matrix and return to the step of normalizing the judgment matrix column by column to obtain the normalized matrix.

[0114] Where i takes values ​​from 1 to n, and n is the order of the judgment matrix. In this embodiment, n can be 3.

[0115] The aforementioned feature parameter weight vector is composed of the initial feature parameter weights corresponding to the three discharge feature parameters, for example, ,in, The initial feature parameter weights corresponding to the discharge type, The initial feature parameter weights corresponding to the discharge frequency are: The initial feature parameter weights corresponding to the average discharge quantity per single instance are determined when the weight values ​​for discharge type, average discharge quantity per single instance, and discharge frequency are 0.668, 0.243, and 0.088, respectively. This determines the feature parameter weight vector. V =[0.668, 0.088, 0.243].

[0116] The above applies to the new vector AV T The i-th component is divided by the i-th component in the transpose of the feature parameter weight vector. To obtain the i-th target ratio, we can take the following example: New vector AV T for Feature parameter weight vector T The first target ratio is The second target ratio is The third target ratio is .

[0117] The formula for calculating the above consistency index can be as follows: ; Where CI is the consistency index and λmax represents the largest eigenvalue.

[0118] The above preset threshold can be set according to requirements, and can be, but is not limited to, a value of 0.1.

[0119] Step S105: Plot the point corresponding to the target's three-dimensional coordinates on the insulation state judgment sphere. Determine the insulation state of the target power distribution equipment based on the area where the point corresponding to the target's three-dimensional coordinates falls.

[0120] In this embodiment, if the point corresponding to the target's three-dimensional coordinates falls into the normal area, the insulation state of the target power distribution network equipment is determined to be at the normal level; if the point corresponding to the target's three-dimensional coordinates falls into the attention area, the insulation state of the target power distribution network equipment is determined to be at the attention level; and if the point corresponding to the target's three-dimensional coordinates falls into the danger area, the insulation state of the target power distribution network equipment is determined to be at the danger level.

[0121] In this embodiment, through steps S101-S105, three discharge characteristic parameters are fused into a target three-dimensional coordinate point, and an insulation status judgment sphere divided into normal, warning, and danger zones is drawn on the computer interface, realizing intuitive visualization of insulation status diagnosis. Maintenance personnel do not need to interpret complex data; they can instantly complete the status assessment simply by observing the area where the target three-dimensional coordinate point is located in the judgment sphere, greatly improving diagnostic efficiency and the intuitiveness of human-computer interaction. Furthermore, by acquiring multiple target discharge characteristic parameters, determining the target three-dimensional coordinates based on these parameters, and then determining the insulation status of the distribution network equipment based on these coordinates, multiple dimensions of discharge characteristic parameters can be comprehensively considered. This overcomes the limitations of traditional single-parameter evaluation, significantly improving the comprehensiveness and accuracy of status assessment, and enabling full utilization and analysis of partial discharge data.

[0122] In some embodiments, after plotting the point corresponding to the target's three-dimensional coordinates on the insulation state determination sphere, the method may further include the following steps: Connecting the points corresponding to the target's three-dimensional coordinates with historical adjacent points yields the connection curve of the target's discharge source.

[0123] When the point corresponding to the target power source gradually moves towards the attention area or danger area in the insulation state judgment ball based on the connection curve of the target power source, the user is reminded to take intervention measures for the distribution network equipment.

[0124] The aforementioned historical adjacent points are those points that are temporally adjacent to the points corresponding to the target's three-dimensional coordinates among the points corresponding to the target's discharge source. Two points that are temporally adjacent among the points corresponding to the target's three-dimensional coordinates are connected.

[0125] To enhance the visualization of the displayed ball, the normal area, attention area, and danger area can be displayed in different colors for easy differentiation. For example, the normal area, attention area, and danger area can be displayed in green, yellow, and red, respectively.

[0126] In this embodiment, the daily comparison of partial discharge fall-in areas can be achieved, thereby enabling the judgment of the trend of changes in the insulation status of distribution network equipment. When the vector (target three-dimensional coordinate point) gradually moves towards the area of ​​concern or danger, it can be regarded as a signal of declining insulation performance, prompting maintenance personnel to take timely intervention measures. Conversely, if the vector remains stable in the normal area for a long period of time, it indicates that the equipment insulation status is good, which can extend the testing and maintenance cycle and reduce maintenance costs. Visualization makes it easier for users to view and improves the efficiency of users in extracting information.

[0127] To better understand the above method, the embodiments of this application provide the following complete embodiments to illustrate the method: The insulation condition of power distribution network equipment directly affects operational safety, reliable power supply, and stable access to distributed energy resources. However, during manufacturing, installation, or operation, power distribution network equipment (such as ring main units, high-voltage switchgear, and cable terminals) often develop latent insulation defects (such as internal air gaps, metal particles, insulation cracks, and surface discharge traces) due to process deviations, environmental stresses, or aging factors. These early insulation defects are difficult to monitor and diagnose. If they are not identified and addressed in a timely manner, they will deteriorate rapidly under long-term electrical, thermal, and mechanical stresses, accompanied by the continuous development of partial discharge (PD) phenomena, gradually expanding into grounding faults or even phase-to-phase short circuits, ultimately leading to unplanned power outages and severely impacting the reliability of power supply to users and the stable operation of the regional energy system.

[0128] Early insulation condition monitoring and diagnosis in distribution networks is challenging. If not addressed promptly, early insulation defects can escalate into grounding and phase-to-phase faults, leading to power outages and impacting power supply reliability and the absorption of distributed energy resources. To address this, some researchers have proposed using partial discharge (PD) monitoring signals to detect early insulation degradation in distribution networks, providing technical support for operation and maintenance. However, the current lack of PD-based health status diagnostic models and algorithms for distribution networks results in the underutilization of a large amount of online or offline PD data, hindering efficient operation and maintenance of distribution networks. Therefore, there is an urgent need to develop a PD-based insulation condition diagnostic model for distribution networks.

[0129] Therefore, how to construct an intelligent insulation condition diagnostic model and method that can effectively integrate multi-dimensional discharge characteristics and possesses good generalization ability and engineering applicability has become a core technical problem that urgently needs to be solved. There is an urgent need to invent a comprehensive evaluation method for the insulation condition of distribution network equipment, providing objective, interpretable, and hierarchical assessment of the insulation condition of the distribution network, and offering key technical support for the health management and risk control of distribution network equipment.

[0130] The lack of an effective mapping model between existing massive online or offline partial discharge data of distribution network equipment and the actual insulation health status of the equipment leads to the underutilization of a large amount of partial discharge data, hindering accurate equipment condition assessment and efficient operation and maintenance decisions. This invention proposes using discharge type, discharge frequency, and average discharge quantity per discharge as feature parameters. A fuzzy evaluation vector weighted average method is used to standardize these feature parameters, eliminating differences in the dimensions and value ranges of each feature. Secondly, to avoid weakening the effect of key features or amplifying the influence of secondary features, this invention uses the analytic hierarchy process (AHP) to weight different feature parameters. Finally, based on the weighted feature parameters, the insulation health status of distribution network equipment is intuitively presented in three-dimensional space as an evaluation sphere. Each monitoring result corresponds to a unique spatial vector in the three-dimensional coordinate system. Its projection on the coordinate axes clearly reflects the actual performance of each key feature, while the vector magnitude serves as a direct quantitative basis for insulation status classification. The monitoring results over multiple days directly reflect the current trend of the equipment's insulation health status. This method effectively integrates multiple feature information, fully reflects the comprehensive role of each indicator, and makes the assessment of insulation health status more scientific, quantifiable, and has good visualization effects.

[0131] This invention proposes a comprehensive evaluation method and system for the insulation status of power distribution network equipment, including the following steps 1-4.

[0132] Step 1: Determination of discharge characteristic parameters of the distribution network insulation state diagnostic model based on partial discharge data.

[0133] This invention comprehensively considers the energy intensity, occurrence frequency, and discharge mechanism of partial discharge signals, and selects three types of discharge characteristic parameters (hereinafter referred to as characteristic parameters) to form a state variable set X = {x1, x2, x3}, where x1 = discharge type, x2 = average discharge quantity per single discharge, which can reflect the energy intensity of local insulation degradation and is an important indicator for judging the severity of defects, and x3 = discharge frequency, which is used to measure the activity level of discharge and helps to identify whether the insulation defect is in a continuous development or stable stage.

[0134] Step 2: Dual-strategy denoising of the distribution network insulation status diagnostic model based on partial discharge data.

[0135] To improve the accuracy of the evaluation model, this invention introduces a specialized discharge data denoising method before extracting state parameters. This includes two strategies: first, eliminating discharge sources that occur less than 5 times to remove the influence of occasional interference pulses on the evaluation results (i.e., eliminating the second discharge source); second, eliminating discharge sources that, although occurring frequently, consistently have a single discharge quantity below 5 pC (i.e., eliminating the third discharge source) to prevent persistent low-energy interference from misleading equipment state judgment. These denoising steps effectively filter out interference components, retaining true and valid partial discharge data, laying a solid foundation for subsequent feature extraction and evaluation.

[0136] Step 3: Establishment of a distribution network insulation status diagnostic model based on partial discharge data.

[0137] Using three characteristic parameters—discharge type, discharge frequency, and average discharge amount per discharge—to establish x, y, and z axes respectively, we establish x 2 +y 2 +z 2 =r 2 An insulation status assessment sphere for power distribution network equipment. Based on the ease of triggering different discharge types at different voltage levels, the discharge types are quantified, with a quantization range of 0-1. Higher values ​​indicate easier punishment and more severe consequences for that discharge type; therefore, different discharge types correspond to different x-axis value ranges.

[0138] In the process of establishing the insulation status judgment sphere for distribution network equipment, in order to eliminate the unreasonable dominance of a certain feature due to its large value caused by differences in the dimensions and value range of the feature parameters, this invention proposes to use the fuzzy judgment vector weighted average method to standardize the feature parameters. The standardized quantization values ​​of discharge type, discharge frequency, and single average discharge quantity are s, respectively. x s y s z The values ​​range from [0, 3]. Secondly, to avoid weakening the role of key features or amplifying the influence of secondary features, it is necessary to weight different feature parameters to more objectively reflect the actual importance of each indicator in the comprehensive evaluation, ensuring that the evaluation results are more consistent with actual operating rules. The feature parameter weight vector was confirmed using the analytic hierarchy process as V = [0.668, 0.243, 0.088]. This feature parameter weight vector includes the preset feature parameter weights corresponding to discharge type, discharge frequency, and average discharge quantity per discharge, i.e., the preset feature parameter weights corresponding to the discharge type. The preset feature parameter weight is 0.668, corresponding to the discharge frequency. The preset characteristic parameter weight is 0.243, corresponding to the average discharge quantity per cycle. It is 0.088.

[0139] The feature parameters [x, y, z] are transformed into [x0, y0, z0] = [s after standardization and weighting. x ×v1, y0=s y ×v2, z0 = s z [×v3], which correspond to the actual values ​​of the x, y, and z axes, respectively, i.e., the three-dimensional coordinates are [x0, y0, z0].

[0140] x0, y0, and z0 constitute vector P. Since the level division vector used in the standardization and quantization process is [1, 2, 3], the magnitude |P| of the vector takes the value in the range [0, 1, 2, 3]. Within the range, when 0≤|P|≤ Then the insulation status level of the distribution network equipment is "normal". The magnitude |P| of the vector takes values ​​in ( , Within the range, that is <|P|≤ Then the insulation condition level of the distribution network equipment is "Caution". The magnitude of the vector |P| takes values ​​in ( , Within the range, that is <|P|≤ If the insulation condition level of the distribution network equipment is "dangerous", then the insulation condition level is "dangerous".

[0141] Step 4: Visualize the trend of insulation status changes in power distribution network equipment.

[0142] The insulation status P1, P2, and P3 of the power distribution network equipment over several consecutive days (e.g., Figure 5 , Figure 6 as well as Figure 9 The monitoring results (r1, r2, r3) are plotted on the same insulation condition judgment sphere. The monitoring results vector P between two adjacent days are connected by arrows (i.e., the point corresponding to the target's three-dimensional coordinates is plotted on the insulation condition judgment sphere displayed on the computer equipment, and the point corresponding to the target's three-dimensional coordinates is connected with the historical adjacent points to obtain the connection curve of the target discharge source). This allows for the comparison of the partial discharge falling area each day, thereby enabling the judgment of the trend of changes in the insulation condition of the distribution network equipment. When vector P gradually moves towards the "attention" or "danger" area, it can be regarded as a signal that the insulation performance has deteriorated, prompting maintenance personnel to take timely intervention measures. Conversely, if vector P remains stable in the "normal" area for a long time, it indicates that the equipment insulation condition is good, which can extend the testing and maintenance cycle and reduce maintenance costs.

[0143] It may also include step 5: confidence verification of the distribution network insulation status diagnostic model based on partial discharge data.

[0144] By verifying whether the discharge frequency or the average discharge amount per discharge increases daily using historical data, the reliability of the distribution network insulation status diagnostic model based on partial discharge data can be determined.

[0145] The following provides specific examples of steps 1-4 above.

[0146] Step 1: Determination of discharge characteristic parameters of the distribution network insulation state diagnostic model based on partial discharge data.

[0147] Discharge characteristic parameters are key inputs to diagnostic models, directly affecting the accuracy and reliability of insulation condition assessments. In practical engineering applications, partial discharge signals typically manifest as a series of pulse sequences with varying amplitudes and wide phase distributions. Therefore, it is essential to select core parameters that effectively reflect the essential characteristics of the discharge to achieve accurate identification of insulation degradation conditions.

[0148] This invention comprehensively considers the energy intensity, occurrence frequency, and discharge mechanism of partial discharge signals, and selects three types of characteristic parameters—average discharge quantity per instance, discharge frequency, and discharge type—to form a state variable set X = {x1, x2, x3}, where x1 = discharge type; x2 = average discharge quantity per instance, which reflects the energy intensity of local insulation degradation and is an important indicator for judging the severity of defects; and x3 = discharge frequency, which measures the activity level of discharge and helps to identify whether insulation defects are in a continuous development or stable stage.

[0149] Step 2: Dual-strategy denoising of the distribution network insulation status diagnostic model based on partial discharge data.

[0150] To improve the accuracy of the evaluation model, this invention introduces a specialized discharge data denoising method before extracting state parameters. This includes two strategies: first, eliminating discharge sources that occur less than 5 times to remove the influence of occasional interference pulses on the evaluation results; second, eliminating discharge sources that, although occurring frequently, consistently have a single discharge magnitude below 5 pC to prevent persistent low-energy interference from misleading equipment status judgments. These denoising steps effectively filter out interference components, retaining true and valid partial discharge data, laying a solid foundation for subsequent feature extraction and evaluation.

[0151] Step 3: Establishment of a distribution network insulation status diagnostic model based on partial discharge data.

[0152] Using the three characteristic parameters of discharge type, discharge frequency, and average discharge amount per discharge (which can also be denoted as average discharge amplitude per discharge) as the x, y, and z axes respectively, an x-axis is established. 2 +y 2 +z 2 =r 2 The insulation status judgment ball of the power distribution network equipment, such as Figure 2 As shown, Figure 2This is a schematic diagram of the insulation status judgment sphere for distribution network equipment provided in an embodiment of the present invention. The magnitude of the vector P composed of three characteristic parameters can directly reflect the insulation status of the distribution network equipment. Based on the operating status of the distribution network equipment and the discharge physics phenomena in the development process of partial discharge, combined with general state programming methods, the insulation status of the distribution network equipment is divided into normal, warning, and dangerous, that is, the evaluation set is L={normal, warning, dangerous}, aiming to achieve a comprehensive evaluation of the insulation status of the distribution network equipment. Figure 2 The normal zone is the area from the origin to the green circle; the caution zone is the area from the green circle to the red circle; and the danger (critical) zone is the area from the red circle to the black circle. The specific steps for creating this zone are as follows: Step 3.1: Preprocessing of state parameter data.

[0153] In the process of establishing a fuzzy comprehensive evaluation model for the insulation status of power distribution network equipment, the status parameters (characteristic parameters) include parameters such as single-time average discharge quantity, discharge frequency, and discharge type. Unlike single-time average discharge quantity and discharge frequency, the discharge type parameter is non-data-quantitative information and therefore needs to be quantified.

[0154] Fault consequence values ​​(hereinafter referred to as consequence values) corresponding to different discharge types are defined, ranging from 0 to 1, with higher values ​​indicating more severe potential consequences. Specifically, the consequence values ​​are set as follows: Normal / External Interference Signals: No risk, will not cause insulation failure, consequence value is set to 0; Point Discharge / Floating Discharge: High initiation voltage, low breakdown probability, consequence value is set to 0.1; Metal Particle Discharge / Internal Discharge: May be triggered at medium voltage, with cumulative damage effects, consequence value is set to 0.7; Surface Discharge: May occur even at lower voltages, easily causing insulation continuity failure, consequence value is set to 0.9. This defines the preset quantitative rules. This setting method fully considers the triggering threshold and hazard level of various discharges, providing a quantitative basis for subsequent insulation condition risk assessment, prioritization, and maintenance strategies.

[0155] Step 3.2: Standardization of feature parameters based on the fuzzy evaluation vector weighted average method.

[0156] Because the three characteristic parameters—average discharge quantity per burst, discharge frequency, and discharge type—differ significantly in terms of dimensions and numerical ranges, directly using the original values ​​for three-dimensional mapping or calculation may lead to a single characteristic having an unreasonable dominance in the evaluation results due to its large value, resulting in an imbalance of information weights and even distortion in visualization. To avoid these problems, all characteristic parameters need to be standardized to unify their dimensions and ranges, ensuring their comparability in multidimensional space.

[0157] To address this, this invention proposes a fuzzy evaluation vector weighted average method to standardize the feature parameters. Specifically, the method involves: standardizing the fuzzy evaluation vector R corresponding to each feature parameter... i =[r i1 r i2 r i3 The product of the rating vector [1,2,3]T and the weighted product is calculated to obtain a unique standardized quantization value s. i The value of i ranges from 1 to 3, which can correspond to x, y, and z respectively, i.e.: (1); In the formula, r ij Represents the characteristic parameter x i For evaluation level L j The degree of membership, j takes values ​​from 1 to 3.

[0158] The specific operating steps are as follows: Step 3.2.1: Confirm the feature parameter x1 and the fuzzy evaluation vector R1 of the discharge type.

[0159] Unlike the average discharge quantity and discharge frequency, discharge type is a discrete variable. Therefore, this invention uses a lookup table method to map different discharge types to corresponding membership vectors.

[0160] As can be seen from step 3.1, tip discharge and floating discharge have a low risk level due to their failure consequence factor of only 0.1, and their fuzzy membership is mainly biased towards "normal"; while metal particle discharge, internal discharge and surface discharge have failure consequence factors of 0.7, 0.7 and 0.9 respectively, and their risk level is high, and their fuzzy membership is more distributed in the "attention" or "danger" level.

[0161] Based on the above risk factors (fault consequence factors), a 5×3 membership matrix was established as shown in the following formula, which corresponds to the membership distribution of the five discharge types at the three state levels (equivalent to the preset membership rules).

[0162] (2); For example, when a tip discharge occurs, its fuzzy evaluation vector R1=[0.7, 0.2, 0.1], that is, the membership degree of the "normal" level is 0.7, indicating that tip discharges are mostly in a normal state, the "caution" level is 0.2, and the "danger" level is 0.1, indicating that it is classified as a dangerous state only in very few cases.

[0163] During levitation discharge, the fuzzy evaluation vector is also R1=[0.7, 0.2, 0.1], indicating that levitation discharge is mainly a normal state, and very few are dangerous states.

[0164] The fuzzy evaluation vector of metal particle discharge is R1 = [0.1, 0.2, 0.7]. Among them, the "normal" level is only 0.1, indicating that most of this type belongs to the dangerous state.

[0165] During internal discharge, its fuzzy evaluation vector is also R1 = [0.1, 0.2, 0.7], indicating that most of this type belongs to the dangerous level.

[0166] The fuzzy evaluation vector of surface discharge is R1 = [0.0, 0.1, 0.9], that is, the membership degree of the "normal" level is 0, and almost all belong to the dangerous level, indicating that this type of discharge has the greatest impact on the insulation safety of the equipment.

[0167] To confirm the fuzzy evaluation vector R1 of the specific discharge type, it is necessary to first confirm the discharge type through the partial discharge pattern recognition algorithm, and then determine the specific fuzzy evaluation vector R1 = [λ1, λ2, λ3] of the discharge type.

[0168] Step 3.2.2: Confirm the fuzzy evaluation vector R2 of the characteristic parameter x2, the single - time average discharge amount.

[0169] To achieve the fuzzy expression of the single - time average discharge amount, the present invention uses the K - Means clustering artificial intelligence algorithm to group and learn the single - time average discharge amount in the historical discharge data.

[0170] The original samples are multiple partial discharge records, and each data contains a single - time average discharge amount. Let the total number of samples be n, and the sample set is expressed as: (3); Among them, d i represents the single - time average discharge amount of the i - th one. To ensure the clustering effect, all samples are uniformly normalized to the interval [0, 1].

[0171] Select K = 3 and use the K - Means artificial intelligence clustering algorithm to group the samples: First, randomly select three centers initially; repeat the iterative sample allocation and center update until convergence; finally, obtain the central values of 3 clusters; for the convenience of constructing the membership function, sort the central values in ascending order, that is, b1 < b2 < b3, where b1 corresponds to the central value of the triangular membership function of the "normal" level; b2 corresponds to the central value of the triangular membership function of the "attention" level; b3 corresponds to the central value of the triangular membership function of the "dangerous" level.

[0172] Triangular membership functions are constructed by taking the cluster center as the vertex of the triangle and setting appropriate left and right support widths, as follows: "Normal" membership function μ1(x): center is b1, range is [a1, c1]; "Attention" membership function μ2(x): center is b2, range is [a2, c2]; "Danger" membership function μ3(x): center is b3, range is [a3, c3]. Figure 3 The trigonometric membership function for the single average discharge amplitude provided in this embodiment of the invention has the following form: (4); The support range is set as follows: (5); Assuming the average discharge amount of a sample is 45pC, we only need to call the trigonometric functions μ1(x), μ2(x), and μ3(x) we just wrote to calculate its membership degree under the three functions "normal", "caution", and "danger", namely the first membership degree, the second membership degree, and the third membership degree.

[0173] Then the fuzzy evaluation vector R2 of the single average discharge quantity is determined as [μ1, μ2, μ3], where μ1, μ2, and μ3 are the first membership degree, the second membership degree, and the third membership degree, respectively.

[0174] Step 3.2.3: Feature parameter x3, fuzzy evaluation vector R3 of discharge frequency is confirmed. The confirmation method is the same as in step 4.2. Finally, the fuzzy evaluation vector R3 of discharge frequency is confirmed as [η1, η2, η3], where η1, η2, and η3 are the first membership degree, the second membership degree, and the third membership degree, respectively. Figure 4 The triangular membership function for the discharge frequency provided in the embodiments of the present invention.

[0175] Finally, the fuzzy evaluation vector R corresponding to each feature parameter is... i = [r i1 r i2 r i3 The value of each feature parameter is calculated by multiplying it with the rating vector [1,2,3]T and performing a weighted product.

[0176] The standardized quantized values ​​of the characteristic parameters discharge type, discharge frequency, and average discharge quantity per discharge are s, respectively. x s y s z The values ​​range from [0, 3].

[0177] Step 3.3: Feature parameter weights v based on the analytic hierarchy process i distribute.

[0178] The influence of characteristics such as discharge type, average discharge quantity per cycle, and discharge frequency on the final evaluation result varies. Simply applying equal weight to each characteristic during the comprehensive evaluation process can easily weaken the role of key characteristics or amplify the influence of secondary characteristics, thus affecting the scientific validity and accuracy of the evaluation results. Therefore, it is necessary to rationally allocate the characteristic parameter weights (i.e., preset characteristic parameter weights, or simply weights) to more objectively reflect the actual importance of each indicator in the comprehensive evaluation, ensuring that the evaluation results better conform to actual operational patterns.

[0179] This invention uses the analytic hierarchy process (AHP) to allocate feature parameter weights. The specific steps are as follows: Step 3.3.1: Construct the judgment matrix.

[0180] Using the Saaty-9 scaling method, the importance of each criterion is compared pairwise to form a judgment matrix A. Elements a in A... ij a represents the importance ratio of element i to element j. ij =1 indicates that element i and element j are equally important; a ij =3 indicates that element i is more important than element j; a ij =5 indicates that element i is more important than element j; a ij =7 indicates that element i is more important than element j; a ij =9 indicates that element i is extremely important compared to element j.

[0181] In this invention, the importance ratio between discharge type and average discharge quantity is assigned a value of 3. The rationale is as follows: Discharge type is the primary basis for determining the nature and risk level of insulation defects. Different types are directly related to equipment failure categories, and their identification plays a decisive role in preventing equipment failure. Average discharge quantity mainly reflects the energy intensity of a discharge event and is important for determining the severity of discharge activity, but it cannot independently reflect the defect type and its development trend. Therefore, a value of 3 is assigned to indicate that discharge type is more important than average discharge quantity, although both have significant reference value in actual diagnosis.

[0182] This invention assigns a value of 7 to the ratio of the importance of discharge type to discharge frequency. The rationale is that the discharge type has a far greater impact on insulation condition than the discharge frequency. Discharge frequency only reflects the activity level of the discharge, not the nature of the defect or its risk level. In practice, even with a high frequency, the harm is limited if the type of discharge is low-risk; conversely, if the discharge type is high-risk, even a low frequency requires serious attention. Therefore, a value of 7 is assigned to reflect the extremely high importance of discharge type compared to frequency.

[0183] In this invention, the importance ratio of average discharge quantity per instance to discharge frequency is assigned a value of 3. The rationale is that average discharge quantity per instance is a key indicator for assessing the energy of a discharge event and its ability to damage insulation. A large discharge quantity often indicates high discharge energy and more severe damage to insulating materials. Discharge frequency, on the other hand, is mainly used to assist in judging the activity level of the discharge and has limited reference value for short-term risk assessment. Therefore, a value of 3 is assigned to reflect that average discharge quantity per instance is more important than discharge frequency, but frequency remains a supplementary parameter that cannot be ignored.

[0184] In summary, the judgment matrix can be obtained as follows: (6); Step 3.3.2: Calculate the weight vector using the summation method.

[0185] Step 1: Normalize by column, i.e., b ij =a ij / Σa ij Therefore, the normalized matrix B is: (7); Step 2: Sum the values ​​of each row of the normalized matrix B. The values ​​of the first to third rows of the normalized matrix B are w1=2.006, w2=0.73, and w3=0.265, respectively.

[0186] Step 3: Normalize again, v i =w i / Σw i The weights of each feature can then be obtained. Therefore, the weight values ​​for discharge type, discharge frequency, and average discharge amount per discharge are 0.668, 0.088, and 0.243, respectively, thus determining the weight vector V = [0.668, 0.088, 0.243].

[0187] Step 3.3.3: Consistency check.

[0188] Step 1: Calculate the maximum eigenvalue λmax using the judgment matrix A and the transpose matrix V of the obtained weight vector. T Multiplying them together yields a new vector AV. T For AV T Divide each component by V T The corresponding component v i We obtain a set of values, and take the average of these values ​​as an approximation of the largest eigenvalue, i.e.: (8); In the formula, n is the order of the judgment matrix.

[0189] Step 2, calculate the consistency index (CI): (9); Step 3: Look up the random consistency index RI in the table. Based on the judgment matrix order n=3, look up the standard table of the analytic hierarchy process to obtain the corresponding RI value (target RI value) of 0.58.

[0190] Step 4: Calculate the consistency ratio (CR): (10); Step 5: Determine consistency. When CR < 0.1, the judgment matrix is ​​considered to have good consistency and the weight results are valid; when CR ≥ 0.1, the judgment matrix needs to be adjusted and the expert scoring needs to be repeated until the consistency requirements are met.

[0191] In summary, the feature parameter weight vector is determined to be V = [0.668, 0.243, 0.088].

[0192] Step 3.4: Visualization of the distribution network insulation status diagnostic model based on partial discharge data.

[0193] like Figure 5 As shown, Figure 5 The schematic diagram provided in this embodiment of the invention shows an insulation state judgment sphere with coordinate points. The standardized and quantized values ​​of the three characteristic parameters for measuring the insulation state correspond to the three coordinate axes of the judgment sphere in three-dimensional space, with the points landing on the x, y, and z axes being x0, y0, and z0, respectively.

[0194] The point where any feature parameter falls on the corresponding coordinate axis is determined by multiplying the standardized feature parameter by its corresponding weight, taking into account that different feature parameters have different degrees of influence on the final evaluation result. That is: x0 = s x ×v1、y0=s y ×v2、z0= s z ×v3, the value range is still within [0, 3].

[0195] x0, y0, and z0 constitute vector P. Since the level division vector is [1, 2, 3], the magnitude |P| of the vector takes values ​​in the range [0, 1, 2, 3]. Within the range of ), i.e., 0 ≤ |P| ≤ Then the insulation status level of the distribution network equipment is "normal". The magnitude |P| of the vector takes values ​​in ( , Within the range, that is ≤|P|< Then the insulation condition level of the distribution network equipment is "Caution". The magnitude of the vector |P| takes values ​​in ( , Within the range, that is ≤|P|≤ If the insulation condition level of the distribution network equipment is "dangerous", then the insulation condition level is "dangerous".

[0196] Using the visualization modeling method described above, the insulation health status of power distribution network equipment can be intuitively presented in three-dimensional space in the form of an evaluation sphere. Each piece of equipment corresponds to a unique spatial vector, and its projection on the coordinate axes clearly reflects the actual performance of each key feature, while the vector magnitude serves as a direct quantitative basis for classifying the insulation status level.

[0197] Step 4: Visualize the trend of insulation status changes in power distribution network equipment.

[0198] like Figure 5 As shown, the insulation status P1, P2, P3, ... (P1, P2, P3, ... represent the points corresponding to the target's three-dimensional coordinates) of a certain device over several consecutive days are plotted on the same insulation status judgment sphere. The monitoring results vector between adjacent days is shown. Connecting the arrows allows for daily comparison of partial discharge fall areas, thereby enabling the determination of trends in the insulation status of distribution network equipment. When the vector... As the vector gradually moves towards the "Caution" or "Critical" zone, it can be seen as a signal of declining insulation performance, prompting maintenance personnel to take timely intervention measures. Conversely, if the vector P remains stable in the "Normal" zone for a long period, it indicates that the equipment insulation is in good condition, which can extend the testing and maintenance cycle and reduce maintenance costs.

[0199] Step 5: Confidence verification of the distribution network insulation status diagnostic model based on partial discharge data.

[0200] By verifying whether the discharge frequency or the average discharge amount per discharge increases daily using historical data, the reliability of the distribution network insulation status diagnostic model based on partial discharge data can be determined.

[0201] Furthermore, the complete embodiments described above are summarized below, along with examples of specific numerical calculations. The operational flow of the comprehensive evaluation method and system for the insulation status of power distribution network equipment proposed in this invention in practical applications is as follows: Figure 6 As shown, Figure 6 A flowchart illustrating the practical operation of the comprehensive evaluation method for insulation status of power distribution network equipment provided in this embodiment of the invention.

[0202] The curve showing the change of the voltage across the impedance detected by the discharge sensor over time is shown below. Figure 7 As shown, Figure 7 This is a schematic diagram of the voltage across the impedance versus time curve provided in an embodiment of the present invention, following the example shown below. Figure 6 The flowchart shown is shown.

[0203] Step 1: Convert the signal voltage amplitude (mV) to apparent discharge quantity (pC) to obtain the curve of apparent discharge quantity changing over time. Different detection methods use different conversion formulas due to their different principles; even for the same detection method, the conversion formula may differ depending on the manufacturer's sensor, design, and manufacturing process. Specific situations require specific analysis.

[0204] The second step involves a dual denoising strategy: removing discharge data with fewer than 5 pulse counts and removing discharge data with a discharge amount less than 5 pC. This strategy aims to preserve accurate and valid partial discharge data, laying a solid foundation for subsequent feature extraction and evaluation. Figure 8 shows a comparison of the PRPS spectra before and after denoising according to an embodiment of the present invention. This denoising strategy effectively removes background noise and atypical discharge interference points, making the retained data more representative and valuable for diagnosis. Specifically, Figure 8(a) shows the PRPS spectrum before denoising according to an embodiment of the present invention, and Figure 8(b) shows the PRPS spectrum after denoising according to an embodiment of the present invention.

[0205] Step 3: Extract characteristic parameters such as discharge type, average discharge quantity per cycle, and discharge frequency based on the denoised PRPS spectrum. The discharge type is internal discharge, the average discharge quantity per cycle is 8.5 pC, and the discharge frequency is 35 cycles.

[0206] Step 4: The fuzzy evaluation vector for the defect discharge type based on the fault consequences is [0.1, 0.2, 0.7]; the fuzzy evaluation vector for determining the average discharge quantity per instance based on K-Means clustering is [0.9845, 0.0155, 0]; and the fuzzy evaluation vector for determining the discharge frequency based on K-Means clustering is [0.7273, 0.2727, 0].

[0207] Step 5: For each feature parameter, obtain its corresponding fuzzy evaluation vector R. i = [r i1 r i2 r i3 After that, it is multiplied by the grade score vector [1, 2, 3]T to obtain the standardized quantization value. Based on this, the standardized quantization value of the discharge type is s. x =2.6, the standardized quantization value of the average discharge quantity per single charge is s z =1.0155, the normalized quantization value of the discharge frequency is s y =1.2727.

[0208] Step 6: To avoid the difference in the influence of different feature parameters on the final evaluation result, the analytic hierarchy process is used to weight the different feature parameters, and then the weighted discharge feature corresponding vector (target three-dimensional coordinates) [x0, y0, z0]=[1.7368, 0.309, 0.089].

[0209] Step 7: Visual modeling of the insulation health status of distribution network equipment. The standardized quantified values ​​of the three characteristic parameters affecting the insulation status (discharge type, discharge frequency, and average discharge quantity per discharge) correspond to the three coordinate axes of the evaluation sphere in three-dimensional space. The points on the x, y, and z axes are x0, y0, and z0, respectively. The magnitude |P| of the vector P=[x0, y0, z0] is 1.766, which is greater than... less than Therefore, the current insulation status of the equipment is L2, which indicates that abnormal signals such as weak discharge have been detected. The equipment can continue to operate under this condition, but it is necessary to strengthen follow-up observation, monitor the operating status in real time, find out the cause of the fault as soon as possible, and shorten the testing and maintenance cycle.

[0210] Step 8: Figure 9 This is a visualization of the continuous multi-day monitoring results of the equipment insulation status provided in this embodiment of the invention. The coordinates of the monitoring results between adjacent days are connected by arrows. The characteristic parameters for the second day are: discharge type is internal discharge, average discharge amount per discharge is 11.0 pC, discharge frequency is 37 times, and the corresponding vector is P2=[1.7368, 0.419, 0.102]; the characteristic parameters for the third day are: discharge type is internal discharge, average discharge amount per discharge is 12.0 pC, discharge frequency is 60 times, and the corresponding vector is P3=[1.7368, 0.521, 0.210]. Therefore, the detection vector magnitudes |P1|, |P2|, and |P3| show an increasing trend over the three consecutive days, indicating that the partial discharge activity of the equipment is gradually increasing, indicating that the insulation deterioration is aggravated, the risk of partial discharge is increased, and key attention and timely maintenance measures are required. The model also demonstrates a reliable confidence level.

[0211] In summary, the comprehensive evaluation method and system for the insulation status of power distribution network equipment proposed in this invention can accurately quantify and dynamically assess the insulation health status of equipment through multi-dimensional characteristic parameters. Through continuous monitoring over multiple days, it can reflect the changing trends of the equipment's insulation status in real time, providing a scientific basis for equipment operation and maintenance management, risk warning, and decision-making.

[0212] This invention achieves several technological breakthroughs in the field of insulation status assessment for power distribution network equipment. First, it innovatively integrates multiple characteristics such as discharge type, discharge frequency, and average discharge quantity per discharge, overcoming the bottlenecks of traditional assessment methods, such as reliance on single characteristics and unscientific weight allocation of indicators. Second, it innovatively introduces the fuzzy evaluation vector weighted average method and the analytic hierarchy process (AHP), achieving standardization of feature parameters and weight allocation, greatly improving the objectivity and accuracy of the assessment results. Furthermore, this invention is the first to map the weighted results of multiple features into three-dimensional space, realizing the visualization and quantitative grading of insulation health status through evaluation sphere vectors. It intuitively demonstrates the comprehensive effect of each key feature, making changes in equipment status readily apparent, facilitating maintenance personnel to intuitively control status evolution trends and health levels, and supporting scientific decision-making.

[0213] To better implement the above method, embodiments of this application provide an apparatus for determining the insulation status of power distribution network equipment. This apparatus may include computer equipment. Figure 10 A structural block diagram of the device for determining the insulation state of power distribution network equipment provided in an embodiment of the present invention is shown below. Figure 10 As shown, the device 100 specifically includes: Module 110 is used to construct a three-dimensional rectangular coordinate system. The three discharge characteristic parameters of the power supply of the power distribution network equipment are mapped to the three-dimensional rectangular coordinate system as x, y, and z axes, respectively, so as to construct a spherical equation based on the discharge characteristic vector.

[0214] The drawing module 111 is used to draw the insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system in the interface of the computer device according to the sphere equation; wherein the insulation status judgment sphere is divided into normal area, attention area and danger area.

[0215] The parameter acquisition module 112 is used to acquire three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period.

[0216] The coordinate determination module 113 is used to determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system based on the three target discharge characteristic parameters.

[0217] The state determination module 114 is used to draw the point corresponding to the three-dimensional coordinates of the target on the insulation state judgment sphere, and determine the insulation state of the target power distribution equipment according to the area where the point corresponding to the three-dimensional coordinates of the target falls.

[0218] The device 100 for determining the insulation status of power distribution network equipment provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0219] Figure 11 An internal structural diagram of a computer device in one embodiment is shown. This computer device can specifically be a terminal or a server. Figure 11 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program causes the processor to perform the steps in the above-described method embodiments. The internal memory may also store a computer program, which, when executed by the processor, causes the processor to perform the steps in the above-described method embodiments. Those skilled in the art will understand that... Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0220] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps in the above method embodiments.

[0221] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, causes the processor to perform the steps in the above method embodiments.

[0222] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0223] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0224] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for determining the insulation status of power distribution network equipment, characterized in that, Applied to a computer device, the method includes: A three-dimensional rectangular coordinate system is constructed, and the three discharge characteristic parameters of the power supply of the power distribution network equipment are mapped to the three-dimensional rectangular coordinate system as x, y, and z axes, respectively, in order to construct a spherical equation based on the discharge characteristic vector; Based on the sphere equation, an insulation status judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system is drawn on the interface of the computer device; wherein, the insulation status judgment sphere is divided into normal area, attention area and danger area. Obtain three target discharge characteristic parameters of the target discharge power source of the target distribution network equipment within a preset time period; Based on the three target discharge characteristic parameters, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system. The point corresponding to the three-dimensional coordinates of the target is plotted on the insulation state judgment sphere. The insulation state of the target power distribution equipment is determined based on the area where the point corresponding to the three-dimensional coordinates of the target falls.

2. The method according to claim 1, characterized in that, The step of drawing the insulation state judgment sphere of the power distribution network equipment in the three-dimensional rectangular coordinate system in the interface of the computer device according to the sphere equation includes: Obtain the modulus range of each insulation state level obtained by classifying the insulation state of the power distribution network equipment; wherein, the multiple insulation state levels include normal level, warning level and dangerous level; the modulus range is the range of the modulus of the three-dimensional coordinates determined by the three discharge characteristic parameters of the first discharge source, and the insulation state level of the first discharge source is the insulation state level corresponding to the modulus range. For each insulation state level, the maximum value in the modulus division range of the insulation state level is used as the radius of the sphere equation to obtain the target sphere equation corresponding to the insulation state level. Based on the target sphere equation corresponding to each insulation state level, the target sphere corresponding to each insulation state level in the three-dimensional rectangular coordinate system is drawn in the interface of the computer device. The sphere range corresponding to the normal level is displayed as the normal area, the sphere range corresponding to the attention level minus the sphere range corresponding to the normal level is displayed as the attention area, and the sphere range corresponding to the danger level minus the sphere range corresponding to the attention level is displayed as the danger area, thus obtaining the insulation state judgment sphere.

3. The method according to claim 1, characterized in that, After plotting the point corresponding to the three-dimensional coordinates of the target on the insulation state judgment sphere, the method includes: Connect the point corresponding to the target's three-dimensional coordinates with its historical adjacent points to obtain the connection curve of the target discharge source; the historical adjacent points are those points that are temporally adjacent to the point corresponding to the target's three-dimensional coordinates among the points corresponding to the historical target's three-dimensional coordinates of the target discharge source, and two temporally adjacent points among the points corresponding to the historical target's three-dimensional coordinates are connected. When the point corresponding to the target discharge power source gradually moves towards the attention area or danger area in the insulation state judgment ball according to the connection curve of the target discharge power source, the user is reminded to take intervention measures for the power distribution network equipment.

4. The method according to claim 2, characterized in that, The step of obtaining the modulus range for each insulation state level obtained by classifying the insulation state of the distribution network equipment includes: Obtain the grade scoring vector corresponding to each insulation state level, obtain the value range of the fuzzy evaluation vector and the value range of the feature parameter weight vector corresponding to the three discharge feature parameters of the first discharge power source, and the fuzzy evaluation vector is used to characterize the degree of membership of the quantized value corresponding to the discharge feature parameter to the normal level, the attention level and the danger level respectively. Multiply the value range of the grade scoring vector corresponding to each insulation state level and the value range of the fuzzy evaluation vector to obtain the first value range of the standardized value corresponding to each insulation state level. Multiply the first value range with the value range of the feature parameter weight vector to obtain the second value range of the coordinate values ​​of each insulation state level in three-dimensional coordinates; The modulus is calculated using the second value range of each insulation state level in three-dimensional coordinates. The modulus value range formed by the modulus of the coordinate values ​​of each insulation state level in three-dimensional coordinates is obtained. The modulus value range corresponding to each insulation state level is used as the modulus division range of each insulation state level.

5. The method according to claim 4, characterized in that, The modulus of the three-dimensional coordinates is calculated using the following formula: ； in, Represents the coordinate values ​​of three-dimensional coordinates. The modulus of the three-dimensional coordinate system.

6. The method according to claim 1, characterized in that, The step of determining the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system based on the three target discharge characteristic parameters includes: The three target discharge characteristic parameters are quantized to obtain the target quantized values ​​corresponding to the three target discharge characteristic parameters; For each target discharge characteristic parameter, the target fuzzy evaluation vector corresponding to the target discharge characteristic parameter is determined based on the target quantization value corresponding to the target discharge characteristic parameter. For each target discharge feature parameter, the product of the target fuzzy evaluation vector corresponding to the target discharge feature parameter, the preset level score vector, and the preset feature parameter weight is determined as the target coordinate value corresponding to the target discharge feature parameter. Based on the target coordinate values, determine the target three-dimensional coordinates of the target discharge source in the three-dimensional rectangular coordinate system.

7. The method according to claim 4 or 6, characterized in that, The method further includes: Obtain the judgment matrix corresponding to the three discharge feature parameters; wherein, the judgment matrix includes the importance ratio between each pair of discharge feature parameters among the three discharge feature parameters; The judgment matrix is ​​column-wise normalized using a first normalization formula to obtain a normalized matrix; wherein, the first normalization formula is: b ij =a ij / Σa ij ; Among them, b ij Let a represent the element in the i-th row and j-th column of the normalized matrix. ij Σa represents the element in the i-th row and j-th column of the judgment matrix. ij This represents the sum of the elements in the j-th column of the judgment matrix; The sum of each row of the normalized matrix is ​​calculated to obtain the first summation value, the second summation value, and the third summation value. The second normalization formula is used to normalize the first summation value, the second summation value, and the third summation value respectively, to obtain the preset feature parameter weights corresponding to the three discharge feature parameters; the second normalization formula is as follows: v i = w i / S w i ; in, v i This represents the weight of the i-th preset feature parameter. w i Σ represents the summation value of the i-th term, where i takes values ​​from 1 to 3. w i This represents the sum of the first summation value, the second summation value, and the third summation value.

8. The method according to claim 5, characterized in that, The step of determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter includes: Based on the target quantization value corresponding to the target discharge characteristic parameter, a first membership degree, a second membership degree, and a third membership degree corresponding to the target quantization value are determined from a preset membership degree rule; wherein, the first membership degree is the membership degree of the target quantization value belonging to the normal level, the second membership degree is the membership degree of the target quantization value belonging to the attention level, and the third membership degree is the membership degree of the target quantization value belonging to the danger level, and the membership degree rule includes the correspondence between different quantization values ​​of different discharge characteristic parameters and membership degrees; Based on the first membership degree, the second membership degree, and the third membership degree, the target fuzzy evaluation vector corresponding to the target discharge feature parameter is determined.

9. The method according to claim 5, characterized in that, The step of determining the target fuzzy evaluation vector corresponding to the target discharge feature parameter based on the target quantization value corresponding to the target discharge feature parameter includes: Obtain the set of historical target discharge feature parameters corresponding to the target discharge feature parameters; wherein, the set of historical target discharge feature parameters includes multiple historical target discharge feature parameters; The K-Means artificial intelligence clustering algorithm was used to process the set of historical target discharge feature parameters to obtain the center values ​​of three clusters; Based on the center values ​​of the three clusters, triangular membership functions are constructed for the target discharge characteristic parameters; the triangular membership functions include normal membership functions, attention membership functions, and danger membership functions; Substituting the target quantized value into the normal membership function yields the first membership degree; substituting the target quantized value into the attention membership function yields the second membership degree; and substituting the target quantized value into the danger membership function yields the third membership degree. Based on the first membership degree, the second membership degree, and the third membership degree, determine the fuzzy evaluation vector corresponding to the target discharge characteristic parameter; The expression for the trigonometric membership function is as follows: ； ； Where x is the target discharge characteristic parameter, μ k μ(x) is the membership function, k is a constant, taking values ​​of 1, 2, and 3 respectively. When k is 1, μ1(x) is the normal membership function; when k is 2, μ2(x) is the attention membership function; when k is 3, μ3(x) is the attention membership function; b1 is the minimum value among the center values ​​of the three clusters; b3 is the maximum value among the center values ​​of the three clusters; b2 is the remaining center value after removing the minimum and maximum values ​​from the center values ​​of the three clusters; and D is the set of historical target discharge characteristic parameters.

10. The method according to claim 1, characterized in that, The acquisition of three target discharge characteristic parameters of the target power source of the target distribution network equipment within a preset time period includes: Acquire a first partial discharge dataset monitored on the power distribution network equipment within a preset time period; wherein the first partial discharge dataset includes partial discharge data of at least one discharge source; The partial discharge data of at least one of the second and third discharge sources are removed from the first partial discharge dataset to obtain the second partial discharge dataset after removal; wherein, the number of discharge events of the second discharge source within the preset time period does not exceed the preset number, and the discharge frequency of the third discharge source exceeds the preset frequency but the single discharge amount is lower than the preset discharge amount. Three target discharge feature parameters of the target discharge source are extracted from the second partial discharge dataset; wherein the target discharge source is any discharge source included in the second partial discharge dataset.

Images