Method for contactless measurement of voltage on a conductor in a corona discharge space charge distortion electric field

Abstract

This invention relates to a non-contact measurement method for conductor voltage based on the space charge distortion electric field of corona discharge, belonging to the field of electromagnetic measurement. By coupling an air corona discharge plasma chemical reaction system with a fluid dynamics model, a plasma-fluid-based needle-plate discharge model was established. The development process of corona discharge and the space charge distribution and electric field changes under different voltage conditions were analyzed, revealing that the discharge can reach a stable state under different voltages. Under this condition, the amplitude of electric field fluctuations within the region is directly related to the voltage applied to the conductor. The reverse voltage obtained by this measurement method has an error of less than 1% compared to the actual voltage, enabling accurate measurement of conductor voltage even considering the space charge distortion of corona discharge. This is of great significance for online monitoring of electrical equipment status and power system protection under high voltage levels or impulse voltage conditions.

Description

Technical Field

[0001] This invention belongs to the field of electromagnetic measurement and relates to a non-contact method for measuring conductor voltage in the electric field of corona discharge space charge distortion. Background Technology

[0002] Currently, voltage measurements on power transmission conductors are divided into two categories: contact measurements and non-contact measurements. Contact voltage measurements, such as those using bushing end screens and RC voltage dividers, require a direct electrical connection between the sensor and the high-voltage equipment in the power system. While offering high measurement accuracy, these methods place extremely high demands on the sensor's insulation performance. Non-contact measurements, such as D-dot and electro-optic effects, do not require direct connection to electrical equipment. They offer advantages such as convenient installation and maintenance, and high safety, but require inversion of the measurement data to obtain the voltage signal. Significant progress has been made in the inversion of power frequency voltage, particularly in addressing the iterative divergence problem during the inverse solution process. However, when defects on the transmission conductor create electrodes with large curvatures, the air near the electrodes will ionize after the voltage rises to a certain level, forming a corona discharge. This generates a large number of charged particles, altering the distribution of spatial charge and causing distortion of the spatial electric field. In the case of space charge distortion electric field coupling interference, if only the nominal electric field generated by the conductor is considered for inversion, the inversion result will differ greatly from the true value. It cannot be well used for non-contact voltage measurement on high voltage level or transmission conductors that generate corona discharge due to impulse voltage.

[0003] In existing methods for measuring conductor voltage, contact-based measurements are problematic for high-voltage applications due to concerns about insulation safety and the unsuitability for live-line operations. Meanwhile, current non-contact methods fail to account for spatial electric field distortion under corona discharge conditions, resulting in inaccurate conductor voltage readings. Therefore, a non-contact method for measuring conductor voltage that considers the distorted electric field caused by corona jet space charge is urgently needed. This would enable accurate non-contact measurement of conductor voltage under space charge interference conditions, thereby improving the safety and reliability of large AC / DC hybrid power grids. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a non-contact method for measuring conductor voltage of corona discharge space charge distortion electric field.

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

[0006] A non-contact method for measuring conductor voltage in the space charge distortion electric field of corona discharge, comprising the following steps:

[0007] S1: Analyze the spatial electric field distortion under different voltages based on the plasma-fluid needle plate discharge model;

[0008] S2: Analyze the spatiotemporal variation of the electric field within the domain;

[0009] S3: Determine the electric field characteristic quantities in the source inversion;

[0010] S4: Conductor voltage inversion based on LSTM neural network to achieve accurate mapping between electric field and voltage.

[0011] Optionally, in S1, a plasma-fluid-based needle-plate discharge model is established by coupling the air corona discharge plasma chemical reaction system with a fluid dynamics model; by changing the needle electrode voltage, simulation of different discharge stages of positive and negative corona is achieved; before and after the corona occurs, the spatial electric field will be significantly distorted due to the presence of space charge, and the spatial electric field will change differently under different conductor voltages, exhibiting a certain regularity; in the presence of space charge, there is a mapping relationship between the synthesized electric field and the conductor voltage.

[0012] Optionally, in S2, when corona discharge occurs, the magnitude of the electric field intensity at different points within the domain is different. The electric field intensity near the electrode will be stronger, and the minimum electric field intensity occurs between the needle and the plate. When the corona discharge is in the periodic discharge stage, the electric field at each point within the domain will also change periodically. When the corona discharge is in the non-periodic discharge stage, the electric field at each point within the domain will not change with time.

[0013] Optionally, in step S3, electric field characteristic quantities are used to describe the electric field situation and establish a mapping relationship between the synthesized electric field and the conductor voltage; the more information contained in the electric field characteristic quantities, the more accurate the mapping relationship will be; based on consideration of spatiotemporal factors, the maximum, minimum and average values ​​of the electric field at four positions on the axis are determined as electric field characteristic quantities.

[0014] Optionally, in step S4, by inputting electric field feature quantities corresponding to different voltages, the neural network's data processing and learning capabilities are utilized to achieve an accurate mapping between electric field and voltage, avoiding the establishment of a complex dynamic correspondence between space charge, space electric field, and conductor voltage.

[0015] The beneficial effects of this invention are as follows: This invention can realize non-contact and accurate measurement of conductor voltage under space charge interference, thereby improving the safety and reliability of AC / DC hybrid power grids.

[0016] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0017] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0018] Figure 1 This is a schematic diagram of a corona discharge needle plate model;

[0019] Figure 2 This is a diagram showing the space charge distribution under corona discharge conditions.

[0020] Figure 3 A comparison diagram of the electric field distribution with and without space charge;

[0021] Figure 4 The distortion of the spatial electric field under different voltages;

[0022] Figure 5 The spatiotemporal variation law of the electric field during corona discharge;

[0023] Figure 6 The training process of an LSTM neural network;

[0024] Figure 7 The results are based on the conductor voltage inversion using an LSTM neural network. Detailed Implementation

[0025] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0026] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0027] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0028] This invention relates to a non-contact method for measuring conductor voltage considering the space charge distortion electric field of corona jets. This method establishes a plasma-fluid-based needle-plate discharge model by coupling an air corona discharge plasma chemical reaction system with a fluid dynamics model. The development process of corona discharge and the distribution of space charge and changes in the electric field under different voltage conditions are analyzed. The results show that the discharge can reach a steady state (the discharge current pulse amplitude remains constant or the discharge current reaches a stable value) under different voltages. In this case, the amplitude of electric field fluctuations within the region is directly related to the voltage applied to the conductor. To better describe the mapping relationship between voltage and electric field, after fully considering the electric field characteristics at different stages of corona discharge, the maximum, minimum, and average values ​​of the electric field collected after the discharge reaches a steady state are determined as electric field characteristic quantities. After determining the field-source relationship, a large amount of electric field data corresponding to different voltages is input into an LSTM neural network, and the model parameters are continuously optimized to make the training results increasingly closer to the actual results, thus forming a non-contact method for measuring conductor voltage considering the space charge distortion electric field of corona jets. Because the electric field distribution at different voltages and discharge stages was carefully analyzed, this method is applicable to non-contact measurement of conductor voltage from the positive and negative corona initiation stages to the glow discharge stages. The final results show that the reverse voltage obtained by this method has an error of less than 1% compared to the true voltage, enabling accurate measurement of conductor voltage even considering the space charge distortion caused by corona discharge. This is of great significance for online monitoring of electrical equipment and power system protection under high voltage levels or impulse voltage conditions.

[0029] This invention proposes a non-contact method for measuring conductor voltage considering the electric field distortion caused by corona discharge space charge, characterized by the following steps:

[0030] S1: Spatial electric field distortion under different voltages based on the plasma-fluid needle-plate discharge model

[0031] To simulate corona discharge in DC transmission projects, a needle-plate discharge model was used. This model couples the air corona discharge plasma chemical reaction system with a fluid dynamics model, enabling a more accurate study of charge distribution during air corona discharge. The model simulates positive and negative corona discharge by changing the voltage polarity on the needle electrode, and by varying the voltage magnitude to simulate the transition from a non-corona discharge stage to a glowing corona discharge stage. Experimental results show that when the voltage reaches the corona initiation voltage, charged particles generated by the plasma chemical reaction begin to exist in space. These charged particles move directionally under the influence of the electric field and aggregate into distinct positive and negative ion clusters, thus altering the spatial electric field distribution and making it significantly different from that before corona discharge. Furthermore, by changing the voltage on the needle tip conductor to induce different stages of discharge, it was found that the nominal electric field generated by the conductor voltage varies in different stages of corona discharge, causing different rates of charged particle generation, migration, and annihilation. This results in differences in the number and distribution of positive and negative ion clusters, thus making the change in the spatial electric field regular with voltage variations.

[0032] A plasma-fluid-based needle-plate discharge model was established by coupling the air corona discharge plasma chemical reaction system with a fluid dynamics model. By changing the needle electrode voltage, simulations of different discharge stages of positive and negative corona were achieved. Analysis revealed that the spatial electric field undergoes significant distortion due to the presence of space charge before and after corona discharge, and the spatial electric field changes differently under different conductor voltages, exhibiting certain regularities. Therefore, it can be concluded that a mapping relationship still exists between the synthesized electric field (field) and the conductor voltage (source) in the presence of space charge.

[0033] S2: Spatiotemporal variation of the electric field within the domain

[0034] When corona discharge is present, the electric field within the domain varies with time and space. Spatially, due to the small radius of curvature of the needle electrode, the electric field is large at this point, and near the needle electrode, the electric field decreases rapidly with increasing distance from the needle electrode. However, near the plate electrode, the electric field strengthens due to the cumulative effect of charged particles. Therefore, the minimum electric field in the corona discharge space occurs between the needle electrode and the plate electrode. Temporally, during the periodic discharge phase of positive / negative corona, the electric field at each point in space also changes periodically. Furthermore, due to the delay in the transmission of field strength changes by space charge, the electric field changes at different points are not synchronous, but the duration of the period is consistent. In the non-periodic discharge phase, the corona current reaches a stable state, and the electric field at each point in space no longer changes over time, only exhibiting numerical differences.

[0035] During corona discharge, the electric field strength varies at different points within the region, increasing towards the electrodes. The minimum electric field strength occurs between the needles and the plate. Furthermore, during the periodic discharge phase, the electric field at each point within the region changes periodically; however, during the non-periodic discharge phase, the electric field at each point within the region remains constant over time.

[0036] S3: Determination of electric field characteristic quantities in source inversion

[0037] To derive the voltage across the conductor from the measured electric field, it is necessary to accurately describe the spatial electric field using electric field characteristic quantities and establish a mapping relationship with the voltage. The analysis above regarding the distribution and variation of the spatial electric field during corona discharge shows that electric field characteristic quantities obtained from both temporal and spatial perspectives are beneficial for accurate description of the electric field. Therefore, the maximum and minimum electric field values ​​at four different distances from the needle pole on the axis are selected as electric field characteristic quantities. Furthermore, considering the differences in electric field waveforms between periodic and non-periodic discharge stages, the average electric field value after the discharge reaches a steady state is also included as one of the electric field characteristic quantities.

[0038] To achieve the inversion from field to source, it is necessary to summarize the electric field situation using electric field characteristic quantities, thereby establishing a mapping relationship between the synthesized electric field and the conductor voltage. In particular, the more information contained in the electric field characteristic quantities, the more accurate the obtained mapping relationship will be. Based on considerations of spatiotemporal factors, this invention determines that the maximum, minimum, and average electric field values ​​at four locations on the axis are used as electric field characteristic quantities.

[0039] S4: Conductor voltage inversion based on LSTM neural network

[0040] Due to the complex and time-varying distribution of space charge, it is difficult to quantitatively analyze voltage and establish a relationship between electric field and voltage. Therefore, a neural network can be used to fit the relationship between electric field and voltage, ultimately retrieving the voltage. This invention uses electric field characteristics obtained under different voltages as independent variables and the corresponding voltages as dependent variables, inputting them into the neural network for training. By adjusting the parameters of the neural network and reducing the training loss value, the model is continuously optimized, making the training results approach the true results, until the maximum number of training iterations is reached. Furthermore, a portion of the electric field-voltage data is reserved as a test set to verify the accuracy of the retrieving results based on this neural network model.

[0041] By inputting electric field characteristics corresponding to different voltages, and utilizing the rapid processing and learning capabilities of LSTM neural networks for large amounts of data, an accurate mapping between electric field and voltage is achieved while avoiding the establishment of a complex dynamic correspondence between space charge, space electric field, and conductor voltage.

[0042] (1) See Figure 1 Corona discharge needle plate model

[0043] Figure 1 A schematic diagram of a two-dimensional axisymmetric needle-plate discharge model used to study the characteristics of corona discharge in air is shown. The discharge domain width is a = 4 mm, the needle-plate gap is d = 5 mm, the needle tip radius of curvature is 62.5 μm, and the current-limiting resistor R is set to 5 kΩ. The temperature is set to 300 K, and the air pressure is 1 atm. Furthermore, the model employs a self-consistent plasma-chemical reaction system capable of accurately simulating the air corona discharge process, involving 18 types of particles and encompassing 57 plasma-chemical reactions and 15 surface reactions. When simulating positive or negative corona discharge or different stages of corona discharge, the voltage applied to the needle electrode can be changed.

[0044] (2) See Figure 2 Charge distribution in space under corona discharge conditions

[0045] When corona discharge occurs, a large number of positive and negative ions and electrons are generated and distributed in a regular manner according to the polarity of the electric field. Taking the application of a -2kV voltage to the needle electrode as an example, a large number of positive ions generated during the discharge process accumulate near the needle electrode, while negative ions move towards the plate electrode under the influence of the electric field, resulting in a wider distribution range of negative ions than positive ions.

[0046] (3) See Figure 3 Comparison of electric field distribution with and without space charge

[0047] Without corona discharge, the electric field strength decreases monotonically with increasing distance from the needle electrode, and the decrease is faster the closer to the needle electrode. When corona discharge occurs, the generated space charge has a certain influence on the electric field. The distribution of the electric field along the axis is as follows: near the needle electrode, the electric field decreases rapidly with increasing distance; while near the plate electrode, the electric field strength increases to some extent with increasing distance. This is because positive ions are concentrated near the needle electrode under the influence of the electric field force, causing distortion of the needle electrode's electric field and greatly enhancing its strength. This effect disappears rapidly with increasing distance from the needle electrode, and the electric field strength decreases rapidly. Negative ions, under the influence of the electric field force, move in clusters in the gaps. Negative ions weaken the electric field near the needle electrode and strengthen the electric field near the plate electrode. Therefore, after a certain distance from the needle electrode, the electric field strength increases as one continues to move away from the needle electrode.

[0048] (4) See Figure 4 Distortion of the electric field in space under different voltages

[0049] As the voltage increases, the discharge between the needle and plate transitions from no corona to periodic discharge, eventually reaching the glow discharge stage. During this process, the electric field strength within the space also changes. Near the plate and needle electrodes, the electric field strength increases with increasing conductor voltage. The only exception is at -4000V, where the electric field strength at the needle electrode is relatively lower than at -2125V. This is because during -2125V corona discharge, the electric field at the needle electrode is obtained when the corona current reaches its maximum. At this point, the distance between the negative ion clusters and the needle electrode is sufficiently far, resulting in minimal impact on the needle tip field strength. However, during -4000V corona discharge, the negative ion clusters remain close to the needle electrode and have a higher density, thus weakening the electric field near the needle electrode.

[0050] (5) See Figure 5 Spatiotemporal variation of electric field during corona discharge

[0051] The three curves in the figure represent the changes in the spatial electric field at different locations on the axis over time at -2kV. It can be seen that during the periodic discharge phase, the electric field intensity at the three points on the axis exhibits a phenomenon of first increasing and then decreasing, with periodic changes. However, the greater the distance from the needle tip, the smaller the difference between the maximum and minimum electric field intensity. It was also found that the electric field intensity at the three locations does not change synchronously, exhibiting a short-term lag. This is because the large amount of space charge near the needle tip delays the transmission of changes in field intensity.

[0052] (6) See Figure 6 Training process of LSTM neural network

[0053] The maximum, minimum, and average electric field values ​​at four locations z = -5, -4, -3, and -2 were selected as electric field characteristic quantities. These, along with the target output (voltage), formed a one-dimensional dataset of length 13. Parametric scanning was used to vary the voltage from -1500V to -4000V, obtaining 500 sets of data under negative corona discharge conditions; similarly, varying the voltage from 1500V to 4000V, obtaining 500 sets of data under positive corona discharge conditions. In both cases, 100 sets of data were used as the test set, and the remaining 400 sets were used as the training set.

[0054] During training, the model parameters (weight coefficients, learner's learning rate, etc.) are first initialized. Then, training collector electric field data is input, and the input information is continuously passed forward. Finally, the voltage value is output. The deviation between the output voltage value and the true voltage value is calculated based on the loss function and used as the loss value. If the maximum number of training iterations (Epoch_num) has not been reached at this point, the optimizer is updated, its learning rate is adjusted, and the weight coefficients are updated in the direction of the maximum gradient of the loss function. Then, the next training iteration is performed, and training collector electric field data is input again for training. Finally, after the maximum number of training iterations is reached, the training ends, and a suitable learning rate and weight coefficients are obtained.

[0055] (7) See Figure 7 Conductor voltage inversion results based on LSTM neural network

[0056] The conductor voltage inversion results show that, regardless of whether the corona discharge is positive or negative, the inverted voltage obtained through the LSTM neural network is very close to the actual voltage. Specifically, under negative corona discharge, the average relative error η between the inverted and actual values ​​is 0.53%; while under positive corona discharge, this relative error is slightly larger, at 0.85%. Since the inversion results under both different corona discharge conditions have only very small inversion errors, this method can achieve accurate measurement of conductor voltage.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for contactless measurement of the voltage of a conductor in a corona discharge space charge distortion electric field, characterized in that: The method includes the following steps: S1: Analyze the spatial electric field distortion under different voltages based on the plasma-fluid needle-plate discharge model; establish a plasma-fluid needle-plate discharge model by coupling the air corona discharge plasma chemical reaction system with a fluid dynamics model; the discharge domain width a = 4 mm, the needle-plate gap d = 5 mm, the needle tip curvature radius is 62.5 μm, the current-limiting resistor R is set to 5 kΩ, the temperature is set to 300 K, and the air pressure is 1 atm; the plasma-fluid needle-plate discharge model adopts a self-consistent plasma chemical reaction system that can accurately simulate the air corona discharge process, involving 18 types of particles and including 57 plasma chemical reactions and 15 surface reactions; by changing the needle electrode voltage, the simulation of different discharge stages of positive and negative corona is achieved; the spatial electric field will be significantly distorted before and after the corona occurs due to the presence of space charge, and the spatial electric field will change differently under different conductor voltages; in the presence of space charge, there is a mapping relationship between the synthesized electric field and the conductor voltage; S2: Analyze the spatiotemporal variation of the electric field within the domain; during corona discharge, the electric field strength varies at different points within the domain, increasing near the electrodes, with the minimum strength occurring between the needles and plates; the space charge generated during corona discharge affects the electric field, and the distribution of the electric field along the axis is as follows: near the needle electrode, the electric field decreases rapidly with increasing distance; near the plate electrode, the electric field strength increases with increasing distance; positive ions, under the influence of the electric field, concentrate near the needle electrode, distorting the electric field and enhancing its intensity. This effect disappears rapidly with increasing distance, and the electric field strength decreases quickly; negative ions, under the influence of the electric field, clump together in the gaps, weakening the electric field near the needle electrode and strengthening it near the plate electrode. In this context, the position of the needle electrode is defined as z=0mm, and the direction from the needle electrode to the plate electrode along the axis of symmetry is the negative z-axis, with the plate electrode position being z=-5mm. Under different voltage conditions, the electric field strength at the needle electrode is lower than that at -2125V when the voltage is -4000V. During the periodic discharge phase, the electric field strength at the three positions with z-coordinates of -0.5mm, -2.5mm, and -5mm on the axis exhibits a phenomenon of first increasing and then decreasing, and changing periodically. As the distance from the needle electrode at these three positions increases from 0.5mm to 2.5mm and 5mm, the difference between the maximum and minimum electric field strength decreases sequentially, and the changes in electric field strength at these three positions are asynchronous, exhibiting a time lag. When the corona discharge is in the non-periodic discharge phase, the electric field at each point within the domain does not change with time. S3: Determine the electric field characteristic quantities in the field source inversion; use the electric field characteristic quantities to describe the electric field situation and establish the mapping relationship between the synthesized electric field and the conductor voltage; based on the consideration of spatiotemporal factors, determine the maximum, minimum and average electric field values ​​at four positions on the axis as electric field characteristic quantities. The z coordinates of the four positions are -5mm, -4mm, -3mm and -2mm, respectively. Select the maximum, minimum and average electric field values ​​at these four positions as electric field characteristic quantities, and add them to the target output, i.e., the voltage, to form a set of one-dimensional data with a length of 13. S4: Conductor voltage inversion based on LSTM neural network to achieve accurate mapping between electric field and voltage; by parametrically scanning, the voltage is varied from -1500V to -4000V, acquiring 500 sets of data under negative corona conditions; the voltage is varied from 1500V to 4000V, acquiring 500 sets of data under positive corona conditions; 100 sets of data are used as the test set in each case, and the remaining 400 sets are used as the training set; by inputting the electric field characteristics corresponding to different voltages, the neural network's data processing and learning capabilities are utilized to achieve accurate mapping between electric field and voltage, avoiding the need to establish a complex dynamic space charge-space... The relationship between the electric field and the conductor voltage is established. During training, the model parameters are first initialized, then the training collector electric field data is input. The input information is continuously passed forward, and finally the voltage value is output. The deviation between the output voltage value and the true voltage value is calculated based on the loss function as the loss value. If the maximum number of training iterations has not been reached, the optimizer is updated, its learning rate is adjusted, and the weight coefficients are updated in the direction of the maximum descent gradient of the loss function. Then, the next training iteration is performed, and the training collector electric field data is input again for training. After the maximum number of training iterations is reached, the training ends, and the learning rate and weight coefficients corresponding to the end of training are obtained.

Images