A method, system and device for calculating the edge field non-uniformity coefficient of a sleeve internal aluminum foil plate, and a storage medium

Abstract

The application discloses a kind of sleeve internal aluminium foil plate edge electric field uneven coefficient calculation method, system, equipment and storage medium, and the present application generates the electrode plate arrangement scheme needed by training set by electrode plate equivalent method, then obtains the electric field uneven coefficient of aluminium foil plate edge under each voltage form by simulation calculation to obtain training set, finally, according to training set, the electric field uneven coefficient of electrode plate in any position under electrode plate equivalent method is obtained by statistical learning method.The present application can be used for optimizing the design of aluminium foil plate arrangement of epoxy core in sleeve under the condition that the profile of electrode plate is determined, so that sleeve product has higher safety and reliability.

Description

Technical Field

[0001] This invention belongs to the field of power equipment design, specifically relating to a method, system, equipment, and storage medium for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside a bushing. Background Technology

[0002] In bushing structure design and insulation performance evaluation, designers use empirical formulas proposed in the last century as the design basis for the epoxy core of the bushing. The partial discharge safety margin of the epoxy core is based on the formula S=E. rm / E r Calculate, where E r E represents the average radial electric field strength between adjacent plates. rm The maximum radial permissible field strength is obtained using the empirical formula E. rm =Kd -0.5 or E rm =K'd -0.55 Calculate (d is the radial distance between adjacent plates). Empirical formula E rm =Kd -0.55 The empirical formula for the initiation voltage of corona discharge in air is extended from the empirical formula E. rm =K'd -0.5 Based on the theoretical analysis of the electric field at the edge of the plate of a parallel plate capacitor, the industry believes that the partial discharge inside the epoxy core occurs at the edge of the plate and is determined by the maximum electric field strength at the edge of the plate and the insulation strength of the material.

[0003] Based on the above empirical formula and industry consensus, the coefficient of electric field non-uniformity at the edge of the aluminum foil electrode inside the bushing must satisfy the mathematical relationship η1=K1d 0.5 Or η2=K2d 0.45 This means that the electric field non-uniformity coefficient is only related to the insulation thickness d, but this differs significantly from the calculation results obtained in the simulation calculation of actual bushing structures considering the shape of the aluminum foil electrode plates. The safety margin assessment results obtained from these empirical formulas have large errors (e.g., electrode plate arrangement...). Figure 6 As shown in Figures 7(a) and 7(b), the non-uniformity coefficient of the electric field at the edge of the electrode is not only related to the insulation thickness d, but also closely related to the voltage form, radial position, and whether it is located at the upper or lower end. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention aims to provide a method, system, device, and storage medium for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the bushing. The present invention can accurately calculate the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the bushing, which is beneficial for accurately assessing the safety margin of the epoxy core.

[0005] The technical solution adopted in this invention is as follows:

[0006] A method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve includes the following steps:

[0007] The electrode arrangement scheme of the bushing is randomly generated; when the electrode arrangement scheme is randomly generated, the electrode spacing is randomly set when the electrode outline, the radial coordinates of the zero layer electrode and the radial coordinates of the last layer electrode are all determined.

[0008] Calculate the electric field non-uniformity coefficient at the edge of each electrode in a randomly generated bushing electrode arrangement scheme;

[0009] Based on the training set, the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted is calculated using a statistical learning method. The training set includes multiple samples, and the input of each sample is the radial coordinates of the electrode in the randomly generated sleeve electrode arrangement scheme and the electrode spacing. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated using the input of that sample. The electrode arrangement scheme to be predicted has the same electrode contour, the same radial coordinates of the zero-layer electrode, and the same radial coordinates of the last layer electrode as the randomly generated sleeve electrode arrangement scheme.

[0010] Preferably, when randomly generating the electrode arrangement scheme, if the electrode outline, the radial coordinates of the zero-layer electrode, and the radial coordinates of the last layer electrode are all determined, the spacing between each layer of electrode and the next layer of electrode is first randomly generated, then the radial coordinates of each electrode are determined according to the randomly generated electrode spacing, and then the upper axial coordinates and lower axial coordinates of each electrode are determined according to the electrode outline and the radial coordinates of each electrode.

[0011] Preferably, the randomly set electrode spacing range is 0.5-3mm, and when randomly generating the electrode arrangement scheme of the sleeve, 10 to 20 sets of electrode arrangement schemes are generated.

[0012] Preferably, when calculating the electric field non-uniformity coefficient at the edge of each electrode in the randomly generated bushing electrode arrangement scheme, the calculation is performed using a bushing electric field simulation model, wherein the electrode arrangement scheme in the bushing electric field simulation model adopts the randomly generated bushing electrode arrangement scheme.

[0013] Preferably, in the bushing electric field simulation model, the shape of the electrode plate is simplified to a rounded rectangle with semicircles at both ends.

[0014] Preferably, the maximum electric field strength at the end of the electrode remains consistent before and after the shape of the electrode is simplified.

[0015] Preferably, the statistical learning method employs a locally weighted regression method.

[0016] The present invention also provides a system for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve, comprising:

[0017] Electrode arrangement scheme generation module: used to randomly generate electrode arrangement schemes for bushings; when randomly generating electrode arrangement schemes, the electrode outline, the radial coordinates of the zero-layer electrode, and the radial coordinates of the last layer electrode are all determined, and the electrode spacing is set randomly.

[0018] First calculation module: used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme of randomly generated bushings;

[0019] The second calculation module is used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted using statistical learning methods based on the training set. The training set includes multiple samples, and the input of each sample is the radial coordinates of the electrode in the randomly generated sleeve electrode arrangement scheme and the electrode spacing. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated using the input of that sample. The electrode arrangement scheme to be predicted has the same electrode contour, the same radial coordinates of the zero-layer electrode, and the same radial coordinates of the last layer electrode as the randomly generated sleeve electrode arrangement scheme.

[0020] The present invention also provides an electronic device, comprising:

[0021] One or more processors;

[0022] A storage device on which one or more programs are stored;

[0023] When the one or more programs are executed by the one or more processors, the one or more processors implement the method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve as described above.

[0024] The present invention also provides a storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve as described above.

[0025] The present invention has the following beneficial effects:

[0026] This invention obtains a training set by randomly generating a bushing electrode arrangement scheme and calculating the electric field non-uniformity coefficient at the edge of each electrode in the randomly generated bushing electrode arrangement scheme. This training set includes multiple samples. The input of each sample is the radial coordinates of the electrodes and the electrode spacing in the randomly generated bushing electrode arrangement scheme. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated using the input of that sample. Based on this training set, the electric field non-uniformity coefficient at the edge of each electrode in any desired electrode arrangement scheme can be calculated using statistical learning methods. This invention utilizes mathematical calculations to fully consider the influence of the bushing structure, including the voltage form, the radial position of the electrodes, and the electrode spacing, on the electric field non-uniformity coefficient at the edge of the electrode. Therefore, the accuracy of the calculated electric field non-uniformity coefficient at the edge of each electrode is significantly improved compared to traditional empirical formulas. Furthermore, this invention only needs to construct a training set once, and can quickly calculate the electric field non-uniformity coefficient at the edge of each electrode in any desired electrode arrangement scheme using statistical learning methods, with a calculation time of only a few seconds, thus exhibiting high computational efficiency. Therefore, the calculation method of this invention can accurately and quickly calculate the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the bushing, which can be used to correct the safety margin calculation and is beneficial to the accurate assessment of the safety margin of the epoxy core. Attached Figure Description

[0027] Figure 1(a) is a schematic diagram of the epoxy impregnated paper sleeve; Figure 1(b) is a longitudinal section of part A in Figure 1(a); Figure 1(c) is an enlarged view of part B in Figure 1(b) (i.e., a schematic diagram of the arrangement of aluminum foil plates inside the epoxy impregnated paper sleeve).

[0028] Figure 2 A schematic diagram of the aluminum foil electrode arrangement inside the epoxy impregnated paper sleeve;

[0029] Figure 3 A schematic diagram of the electric field distortion at the end of the epoxy-impregnated paper electrode plate;

[0030] Figure 4 This is a simplified diagram of the end shape of the aluminum foil electrode plate in an embodiment of the present invention;

[0031] Figure 5(a1) shows the distribution of the electric field non-uniformity coefficient at the upper end of the aluminum foil plate under AC voltage and the training set sample data in an embodiment of the present invention; Figure 5(a2) shows the distribution of the electric field non-uniformity coefficient at the lower end of the aluminum foil plate under AC voltage and the training set sample data in an embodiment of the present invention; Figure 5(b1) shows the distribution of the electric field non-uniformity coefficient at the upper end of the aluminum foil plate under DC voltage and the training set sample data in an embodiment of the present invention; Figure 5(b2) shows the distribution of the electric field non-uniformity coefficient at the lower end of the aluminum foil plate under DC voltage and the training set sample data in an embodiment of the present invention.

[0032] Figure 6This refers to the radial spacing distribution of adjacent electrodes in the aluminum foil electrode arrangement scheme used in this embodiment of the invention.

[0033] Figure 7(a) is a comparison between the calculated and simulated results of the electric field non-uniformity coefficient at the end of the aluminum foil electrode plate under AC voltage in the embodiment of the present invention; Figure 7(b) is a comparison between the calculated and simulated results of the electric field non-uniformity coefficient at the end of the aluminum foil electrode plate under DC voltage.

[0034] Figure 8 This is a structural block diagram of the system for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve of the present invention.

[0035] In the figure, 1-epoxy impregnated paper sleeve, 2-electrode plate. Detailed Implementation

[0036] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0037] like Figure 1(a)-Figure 1(c) As shown, a certain number of aluminum foil electrodes are arranged inside the epoxy-impregnated paper sleeve. The electrodes are aligned with the axial direction (…). Figure 1(a)-Figure 1(c) Parallel in the vertical direction, and in the radial direction ( Figure 1(a)-Figure 1(c) The plates are arranged at regular intervals in the left and right directions, and the length of the plates from the inside to the outside is ( Figure 1(a)-Figure 1(c) The number of plates gradually decreases in the vertical direction. The innermost plate is called the zero-layer plate, the outermost plate is called the last-layer plate, and the (i+1)th plate is called the i-th layer plate, where i is a positive integer.

[0038] like Figure 1(a)-Figure 2 As shown, for a design scheme with n+1 plates, the radial positions of the plates from the inside out are defined as r0, r1, ..., r2. n The upper axis coordinate is z up,0 z up,1 ... z up,n The lower axis coordinate is z down,0 z down,1 ... z down,n The difference in radial coordinates between adjacent plates is called the plate spacing. The spacing Δr between adjacent i-th and (i-1)-th plates is... i =r i -r i-1 The difference in axial coordinates between adjacent plates is called the upper step, Δz. up,i =z up,i -z up,i-1 The difference in the lower axial coordinates of adjacent plates is called the lower step, Δz. down,i =z down,i -z down,i-1 .

[0039] like Figure 2 As shown, the electrode profile is defined as the envelope formed by the upper and lower ends of the electrode plate. The envelope at the upper end is the upper profile, and the envelope at the lower end is the lower profile. The upper and lower profiles are represented by the function z. up (r) and z down (r) indicates that the radial coordinate r of the electrode plate is located between the radial coordinate r0 of the zero-layer electrode plate and the radial coordinate r of the last-layer electrode plate. n Between, i.e., r∈[r0,r n In this invention, in the electrode profile z... up (r) and z down Given that (r) the radial position of the electrode is r i Then the axial and lower axial coordinates on the plate satisfy z up,i =z up (r i ) and z down,i =z down (r i ).

[0040] The electric field non-uniformity coefficient is used in high-voltage insulation to determine the degree of electric field non-uniformity; it is defined as the ratio of the maximum electric field strength to the average electric field strength. For the internal electrode plates of the bushing described in this invention, if the maximum electric field strength on the upper surface of the i-th electrode plate is E... max,up,i The maximum electric field strength on the lower end surface of the i-th electrode layer is E. max,down,i The average radial electric field strength between the i-th electrode layer and the (i-1)-th electrode layer is E avg,i Then the electric field non-uniformity coefficient at the upper end of the i-th electrode plate is E. max,up,i / E avg,i The electric field non-uniformity coefficient at the lower end of the i-th electrode layer is E. max,down,i / E avg,i Taking the upper end of the electrode plate as an example, such as... Figure 3 As shown, the electric field is severely distorted near the upper end, while it is very uniform in the region far from the upper end. The degree of distortion of the electric field at the edge of the electrode is closely related to the safety and reliability of the epoxy core.

[0041] This invention generates the electrode arrangement scheme required for the training set through the electrode equivalence method, then obtains the electric field non-uniformity coefficient at the edge of the aluminum foil electrode under various voltage conditions through simulation calculation to obtain the training set, and finally obtains the electric field non-uniformity coefficient of the electrode at any position under the electrode equivalence method based on the training set and through statistical learning method.

[0042] Specifically, the method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve of the present invention includes the following steps:

[0043] Step 1: Randomly generate a series of electrode arrangement schemes

[0044] Step 1.1: Given a determined electrode profile, zero-layer electrode position, and last-layer electrode position, randomly generate electrode spacing Δr1,…,Δr n Ensure Δr1 + ... + Δr n =r n -r0 and Δr i Within the design range (typically 0.5mm to 3mm).

[0045] Step 1.2, based on the geometric relationship Δr i =r i -r i-1 Obtain the radial coordinates of each electrode layer. Since the radial coordinate r0 of the zero-layer electrode is already determined, the formula Δr is used. i =r i -r i-1 It can be seen that the radial coordinates of the first layer of plates are r1 = r0 + Δr1, the radial coordinates of the second layer of plates are r2 = r1 + Δr1 = r0 + Δr1 + Δr2, and so on, until the coordinates of the i-th layer of plates are r1 = r0 + Δr1 + Δr2. i =r i-1 +Δr i = r0 + Δr1 + ... + Δr i In this way, the radial coordinates of each electrode layer can be obtained.

[0046] Step 1.3: Determine the upper and lower axial coordinates of the electrodes based on their radial coordinates and profiles. According to the definition in the conceptual introduction, the upper axial coordinate z... up,i =z up (r i The lower axis coordinate z down,i =z down (r i ).

[0047] Generally, the training set obtained by generating 10 to 20 electrode arrangement schemes can meet the training requirements.

[0048] Step 2: Simulation calculation to obtain the data required for the training set.

[0049] Step 2.1: Simplify the electrode shape to facilitate modeling, reduce the number of meshes, and improve computational efficiency. The end shapes of the electrodes before and after simplification are as follows: Figure 4 As shown. Before simplification, according to existing processing technology, the edge of the electrode plate needs to undergo secondary folding to eliminate burrs on the edge of the electrode plate, and the thickness of the electrode plate varies at different axial positions. After simplification, the electrode plate is approximated as a rounded rectangle with semicircles at both the top and bottom. The geometry of this electrode plate is simpler, and under the same voltage conditions, the maximum electric field strength at the end of the simplified electrode plate is the same as the maximum electric field strength at the end of the actual electrode plate.

[0050] Step 2.2: Construct a bushing electric field simulation model based on the simplified electrode edge morphology, and calculate the non-uniformity coefficient of the electric field at the edge of each electrode using the bushing electric field simulation model.

[0051] Step 2.3: Using the radial coordinates and spacing between one of the electrodes as input samples, and the electric field non-uniformity coefficients at the upper and lower ends of that electrode as output samples, summarize the data from all electrodes to obtain the training set, such as... Figures 5(a1)-5(b2) As shown. Specifically, the training set includes multiple samples. The input of each sample is the radial coordinates of the plates and the distance between the plates in a randomly generated plate arrangement scheme of the bushing. The output of each sample is the non-uniformity coefficient of the electric field at the edge of the plate, calculated from the input of that sample.

[0052] Step 3: Based on the training set, use statistical learning methods to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted;

[0053] Step 3.1: Divide the training set: Divide the training set into four parts according to each voltage form (including DC voltage form and AC voltage form), the upper end of the plate, and the lower end of the plate. That is, the DC voltage form corresponds to the upper end training set and the lower end training set, and the AC voltage form corresponds to the upper end training set and the lower end training set, such as... Figures 5(a1)-5(b2) These four point maps correspond to these four parts of the training set.

[0054] Step 3.2: Given the electrode arrangement scheme to be calculated (predicted) (i.e. the electrode arrangement scheme to be predicted mentioned above), the electrode outline, zero-layer electrode position, and last-layer electrode position of the scheme must be consistent with the electrode outline, zero-layer electrode position, and last-layer electrode position mentioned in Step 1.

[0055] Step 3.3: Based on the geometric relationship Δr i =r i -r i-1 Calculate the radial coordinates and spacing of each electrode in the electrode arrangement scheme to be predicted, and summarize the data.

[0056] Step 3.4: Using the training set from Step 3.1 and the data obtained from Step 3.3, calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted.

[0057] Example

[0058] Under the premise that the electrode outline, the position of the zero-layer electrode, and the position of the last-layer electrode are consistent, with Figures 5(a1)-5(b2) The data shown is the training set, and a locally weighted regression method is used to predict... Figure 6 The electrode spacing distribution shown corresponds to the electric field non-uniformity coefficient at the ends of each electrode in the electrode arrangement scheme.

[0059] Let r be the input of the training set samples mentioned in steps one and two. (j) and Δr (j) The (j) in the upper right corner represents the j-th sample. Since the locally weighted regression method requires separate model training for each input to be predicted, the discussion can be simplified to a scenario where the output is predicted for a given input r and Δr. The specific steps are as follows:

[0060] Step 1: During training, for sample data far from the electrode, their impact on the prediction results can be appropriately reduced; for sample data close to the electrode, their impact on the prediction results should be preserved as much as possible. In the locally weighted review method, the strength of this impact is represented by weights, as shown by w(j) = exp(-||r (j) -r||2 2 / 2τ 2 ), where τ is an adjustable hyperparameter.

[0061] Step 2: Process the electrode spacing and electric field non-uniformity coefficient, and use the mapping result obtained from the bidirectional mapping as the input and output of the local weighted regression method. Define the input of the local weighted regression method as x = (1, Δr) and the output as y = lnη. Perform the same processing on the sample data in the training set to obtain x. (j) =(1,Δr) (j) ), y (j) =lnη (j) The (j) in the upper right corner represents the j-th sample.

[0062] Step 3: Train the locally weighted regression method, i.e., solve the minimization problem. The specific implementation method involves iterative solving, repeatedly calculating the iterative formula. Until convergence is achieved.

[0063] Step 4: Calculate the non-uniformity coefficient of the electric field at the end of the plate, we have lnη=y=θ T x.

[0064] Step 5: Select the voltage type and specify the electric field non-uniformity coefficient as the upper or lower end. For the electric field non-uniformity coefficient corresponding to a certain input r and Δr to be predicted, calculate it sequentially using the methods described in steps 1-4. The electric field non-uniformity coefficient at the edge of the electrode obtained in this way is as follows: Figure 6 As shown, the prediction results of the trained model are very close to the simulation results, and the dark and light points in the figure basically overlap.

[0065] like Figure 8As shown, the present invention also provides a system for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the above-mentioned sleeve, comprising: an electrode plate arrangement scheme generation module for randomly generating an electrode plate arrangement scheme for the sleeve; when randomly generating the electrode plate arrangement scheme, under the condition that the electrode plate outline, the radial coordinate of the zero layer electrode plate and the radial coordinate of the last layer electrode plate are all determined, the electrode plate spacing is randomly set.

[0066] First calculation module: used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme of randomly generated bushings;

[0067] The second calculation module is used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted using statistical learning methods based on the training set. The training set includes multiple samples, and the input of each sample is the radial coordinates of the electrode in the randomly generated sleeve electrode arrangement scheme and the electrode spacing. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated using the input of that sample. The electrode arrangement scheme to be predicted has the same electrode contour, the same radial coordinates of the zero-layer electrode, and the same radial coordinates of the last layer electrode as the randomly generated sleeve electrode arrangement scheme.

[0068] The embodiments of the present invention also provide corresponding devices and computer-readable storage media for implementing the solutions provided in the embodiments of the present invention.

[0069] The device includes a memory and a processor. The memory is used to store instructions or code, and the processor is used to execute the instructions or code to enable the device to perform the redundant card data transmission method described in any embodiment of this application.

[0070] In practical applications, the computer-readable storage medium can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium.

[0071] In bushing electrode design, the electrode profile determines the effectiveness of the uniform electric field within the bushing. Given a fixed electrode profile, the electrode spacing, or the specific arrangement of the electrodes, determines the internal safety and reliability of the epoxy core. When determining the electrode spacing, designers need to calculate the initial partial discharge margin (i.e., the ratio of material insulation strength to the maximum electric field strength at the end of that electrode layer) for each layer. Since the radial electric field distribution of the electrodes is fixed given the electrode profile and voltage conditions, calculating the maximum electric field strength at the electrode ends only requires obtaining the electric field non-uniformity coefficient at the electrode ends. The method provided by this invention allows for the rapid and accurate acquisition of the electric field non-uniformity coefficient at the electrode ends, thereby enabling the calculation of the initial partial discharge margin within the epoxy core.

[0072] Compared to existing empirical formulas, the method proposed in this invention significantly improves accuracy, as empirical formulas often suffer from serious errors. Furthermore, compared to directly obtaining the initial partial discharge margin through simulation calculations, the method proposed in this invention offers significantly higher computational efficiency. When calculating the initial partial discharge margin for a specific electrode arrangement, the former takes approximately 10 minutes, while the latter, with a suitable training set, only requires a few seconds.

[0073] When iteratively optimizing a specific electrode arrangement scheme with a defined electrode profile, the method proposed in this invention can quickly and accurately calculate the initial partial discharge margin of each iteration scheme, thereby enabling rapid and efficient design optimization.

[0074] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for calculating the electric field non-uniformity coefficient at the edge of an aluminum foil electrode plate inside a sleeve, characterized in that, The process includes the following: The electrode arrangement scheme of the bushing is randomly generated; when the electrode arrangement scheme is randomly generated, the electrode spacing is randomly set when the electrode outline, the radial coordinates of the zero layer electrode and the radial coordinates of the last layer electrode are all determined. Calculate the electric field non-uniformity coefficient at the edge of each electrode in a randomly generated bushing electrode arrangement scheme; Based on the training set, the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted is calculated using a statistical learning method. The training set includes multiple samples, and the input of each sample is the radial coordinates of the electrode and the electrode spacing in the randomly generated electrode arrangement scheme of the sleeve. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated using the input of that sample. The electrode arrangement scheme to be predicted has the same electrode outline, the same radial coordinate of the zero-layer electrode, and the same radial coordinate of the last-layer electrode as the electrode arrangement scheme of the randomly generated sleeve; the statistical learning method adopts the local weighted regression method.

2. The method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve according to claim 1, characterized in that, When randomly generating the electrode arrangement scheme, given that the electrode outline, the radial coordinates of the zero-layer electrode, and the radial coordinates of the last layer electrode are all determined, the spacing between each layer of electrode and the next layer of electrode is first randomly generated. Then, the radial coordinates of each electrode are determined based on the randomly generated electrode spacing. Finally, the upper and lower axial coordinates of each electrode are determined based on the electrode outline and the radial coordinates of each electrode.

3. The method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve according to claim 1, characterized in that, The randomly set electrode spacing ranges from 0.5 to 3 mm. When randomly generating the electrode arrangement scheme for the bushing, 10 to 20 sets of electrode arrangement schemes are generated.

4. The method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve according to claim 1, characterized in that, When calculating the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme of the randomly generated bushing, the calculation is performed through the bushing electric field simulation model, and the electrode arrangement scheme in the bushing electric field simulation model adopts the electrode arrangement scheme of the randomly generated bushing.

5. The method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve according to claim 4, characterized in that, In the sleeve electric field simulation model, the shape of the electrode plate is simplified to a rounded rectangle with semicircles at both ends.

6. The method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve according to claim 5, characterized in that, Before and after the shape of the electrode is simplified, the maximum electric field strength at the end of the electrode remains the same.

7. A system for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside a sleeve, characterized in that, include: Electrode arrangement scheme generation module: used to randomly generate electrode arrangement schemes for bushings; When randomly generating the electrode arrangement scheme, the electrode spacing is randomly set when the electrode outline, the radial coordinates of the zero-layer electrode and the radial coordinates of the last layer electrode are all determined. First calculation module: used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme of randomly generated bushings; The second calculation module is used to calculate the electric field non-uniformity coefficient at the edge of each electrode in the electrode arrangement scheme to be predicted, based on the training set and through statistical learning methods. The training set includes multiple samples, and the input of each sample is the radial coordinates of the electrode in the electrode arrangement scheme of the sleeve and the electrode spacing. The output of each sample is the electric field non-uniformity coefficient at the edge of the electrode calculated through the input of that sample. The electrode arrangement scheme to be predicted has the same electrode outline, the same radial coordinate of the zero-layer electrode, and the same radial coordinate of the last-layer electrode as the electrode arrangement scheme of the randomly generated sleeve; the statistical learning method adopts the local weighted regression method.

8. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, It stores a computer program, wherein when the computer program is executed by a processor, it implements the method for calculating the non-uniformity coefficient of the electric field at the edge of the aluminum foil electrode plate inside the sleeve as described in any one of claims 1 to 6.

Images