A Multi-parameter Optimization Method and System for a High-Voltage Resistor Divider

By optimizing the compensation parameters of the high-voltage resistor divider using the dung beetle optimization algorithm, constructing an error model and defining a fitness function, the measurement error problem of the high-voltage resistor divider over a wide frequency and temperature range is solved, achieving high-precision and stable voltage division measurement.

CN123116784APending Publication Date: 2026-11-17CHINA SHIP DEV & DESIGN CENT
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611215683.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-12
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The measurement accuracy of existing high-voltage resistor dividers is affected by frequency and temperature variations over a wide frequency range and temperature range. Existing compensation parameter optimization designs suffer from insufficient global optimization capability and low convergence efficiency, making it difficult to achieve coordinated global optimal matching of multiple compensation parameters, resulting in large measurement errors.

Method used

The compensation parameters of the high-voltage resistor divider are optimized using the dung beetle optimization algorithm. A stray capacitance and temperature error model is constructed, and a fitness function is defined with the voltage division ratio error as the optimization objective. The optimal compensation parameters are output through iterative optimization using the dung beetle optimization algorithm.

Benefits of technology

The system can quickly lock in the optimal frequency and temperature compensation parameters under complex operating conditions, which significantly improves the measurement accuracy and stability of the high-voltage resistor divider and reduces the impact of stray capacitance current fluctuations and temperature changes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123116784A_ABST
    Figure CN123116784A_ABST
Patent Text Reader

Abstract

This invention discloses a multi-parameter optimization method and system for a high-voltage resistor divider. The method includes: constructing a stray capacitance parameter error model in a wide frequency domain based on the distributed capacitance in the high-voltage resistor divider; constructing an operating temperature parameter error model based on the relationship between the voltage divider resistance and temperature; initializing a dung beetle optimization algorithm search population using the compensation parameters of the high-voltage resistor divider as optimization variables, where each individual corresponds to a set of candidate values ​​for the compensation parameters; defining a fitness function with voltage division ratio error as the optimization objective based on the stray capacitance parameter error model and the operating temperature parameter error model; iteratively optimizing the search population using the dung beetle optimization algorithm, and outputting the compensation parameters corresponding to the individual with the best fitness as the optimal compensation parameters for the high-voltage resistor divider. This invention can suppress errors caused by frequency and temperature, and improve the accuracy of voltage division measurement.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of multi-parameter optimization technology for high-voltage resistor dividers, and in particular to a multi-parameter optimization method and system for high-voltage resistor dividers. Background Technology

[0002] Resistive voltage dividers, based on the voltage division principle of series precision resistors, measure high-voltage signal attenuation. They feature simple structure, mature manufacturing process, small size, light weight, and good long-term stability, making them widely used voltage measurement devices in power systems and high-voltage testing. The main body of a high-voltage resistive voltage divider consists of a high-voltage arm resistor and a low-voltage arm resistor connected in series, relying on a fixed resistance ratio to output a low-voltage measurement signal with a corresponding amplitude.

[0003] In practical applications, inherent distributed capacitance exists between the voltage divider resistor and its shield, as well as between different segments of the resistor itself. Furthermore, constraints imposed by insulation structure, heat dissipation design, shielding configuration, and the miniaturization and integration of equipment mean that ambient temperature fluctuations can cause changes in the current characteristics of stray capacitance. Additionally, changes in the frequency of the measured signal can affect the impedance performance of the distributed capacitance. These factors can cause the actual voltage division ratio of the voltage divider to deviate from the theoretical value, resulting in measurement errors. Moreover, the resistance values ​​of the high-voltage and low-voltage arms will experience temperature drift with ambient temperature, further expanding the deviation range of the voltage divider measurement.

[0004] To improve measurement accuracy, existing technologies generally compensate for and correct errors caused by stray capacitance and temperature drift by setting frequency compensation parameters and temperature compensation parameters. However, current compensation parameter optimization designs mostly rely on traditional optimization algorithms, which have shortcomings such as insufficient global optimization capability, low convergence efficiency, and susceptibility to getting trapped in local optima. They are difficult to achieve coordinated global optimal matching of multiple compensation parameters over a wide frequency range and a wide temperature range, and cannot effectively suppress voltage division errors under complex operating conditions, making it difficult to meet the ever-increasing accuracy requirements of high-voltage measurement scenarios. Summary of the Invention

[0005] The main objective of this invention is to optimize the compensation parameters of a high-voltage resistor divider based on the dung beetle algorithm, suppress errors caused by frequency and temperature, and improve the accuracy of voltage divider measurement.

[0006] The technical solution adopted in this invention is: a multi-parameter optimization method for a high-voltage resistor divider, comprising: Based on the distributed capacitance in the high-voltage resistor divider, a stray capacitance parameter error model is constructed in the wide frequency domain; based on the relationship between the voltage divider resistance and temperature, an operating temperature parameter error model is constructed. Using the compensation parameters of the high-voltage resistor divider as optimization variables, the search population of the dung beetle optimization algorithm is initialized. The position information of each dung beetle in the search population corresponds to a set of candidate values ​​of the compensation parameters. Based on the stray capacitance parameter error model and the operating temperature parameter error model, a fitness function with the voltage division ratio error as the optimization objective is defined. The dung beetle optimization algorithm is used to iteratively optimize the search population and output the compensation parameters corresponding to the individual with the best fitness, which are then used as the optimal compensation parameters for the high-voltage resistor divider.

[0007] According to the above technical solution, the compensation parameters specifically include frequency compensation coefficient and temperature compensation coefficient.

[0008] According to the above technical solution, the stray capacitance parameter error model is determined based on the equivalent total distributed capacitance of the voltage divider resistor to ground, the distributed capacitance of the high-voltage end lead to the voltage divider body, the signal frequency, and the total resistance of the high and low voltage arms.

[0009] According to the above technical solution, the determination of the fitness function includes: calculating the error value of the high-voltage resistor divider at a given frequency and temperature; the error value is the absolute value of the sum of the output of the stray capacitance parameter error model and the output of the temperature parameter error model, minus the absolute value of the compensation term calculated by the frequency compensation coefficient, the temperature compensation coefficient and the current frequency and temperature; The maximum value among all error values ​​corresponding to all frequency points within the entire preset working frequency range is taken as the fitness value of the dung beetle algorithm.

[0010] According to the above technical solution, the iterative optimization of the search population using the dung beetle optimization algorithm includes: The initialized dung beetle population was randomly divided into four categories: rolling dung beetle, hatching dung beetle, foraging dung beetle, and thieving dung beetle. In each iteration, the fitness of each type of individual is calculated, and the position update is performed according to the corresponding position update method. Repeatedly calculate fitness and position updates until the preset maximum number of iterations is reached, then stop and output the compensation parameters corresponding to the individual with the best fitness.

[0011] According to the above technical solution, the position update of the dung beetle includes the position update of the unobstructed dung beetle and the position update of the obstructed dung beetle. The next position of the unobstructed dung beetle is determined by the current position, the previous position, the worst-case position, the deflection coefficient, the intensity coefficient, and a random number with a value of -1 or 1. The worst-case position is used to simulate the distance measurement of light intensity changes. The next position of the obstructed dung beetle is determined by the current position, the previous position, and the deflection angle. The position is not updated when the deflection angle is 0, π / 2, or π. The position update method of the dung beetle is as follows: the position at the next moment is determined by the current position, two independent random vectors, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound, multiplied by each random vector, summed, and then added to the current position. The position update method of the foraging dung beetle is as follows: the position at the next moment is determined by the current position, two random vectors that follow a normal distribution, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound respectively, multiplied by each normal random vector, summed, and then added to the current position. The position update method of the dung beetle is as follows: the position at the next moment is determined by the global best position, the distance from the current position to the global best position, a random vector that follows a normal distribution, and a preset constant. The constant is multiplied by the normal random vector and the distance and then added to the global best position.

[0012] According to the above technical solution, the upper and lower boundaries of the optimal foraging area are determined in the following way: The calculation is performed based on the global optimal position, dynamic reduction factor, and preset upper and lower bounds of the search space. The lower bound of the optimal foraging area is the larger of the difference between the global best position multiplied by 1 and the dynamic reduction factor, and the preset lower bound of the search space. The upper bound of the optimal foraging area is the smaller of the sum of the global best position multiplied by 1 and the dynamic reduction factor, and the preset upper bound of the search space. The dynamic reduction factor decreases linearly from 1 to 0 with the number of iterations.

[0013] According to the above technical solution, when initializing the dung beetle population, the location information of each individual dung beetle contains multiple candidate values, which correspond to the frequency compensation coefficient and the temperature compensation coefficient, respectively.

[0014] Another aspect of the present invention provides a multi-parameter optimization system for a high-voltage resistor divider, comprising: The error modeling module is used to construct a stray capacitance parameter error model in a wide frequency domain based on the distributed capacitance in the high-voltage resistor divider; and to construct an operating temperature parameter error model based on the relationship between the voltage divider resistance and temperature. The population initialization module is used to initialize the search population of the dung beetle optimization algorithm with the compensation parameters of the high voltage resistor divider as optimization variables. The position information of each dung beetle in the search population corresponds to a set of candidate values ​​of the compensation parameters. The fitness definition module is used to define a fitness function with the voltage division ratio error as the optimization objective based on the stray capacitance parameter error model and the operating temperature parameter error model. The iterative optimization module is used to iteratively optimize the search population using the dung beetle optimization algorithm, and output the compensation parameters corresponding to the individual with the best fitness as the optimal compensation parameters of the high-voltage resistor divider.

[0015] The beneficial effects of this invention are as follows: By establishing stray capacitance parameter error models and operating temperature parameter error models in the wide frequency domain, the voltage division ratio deviation caused by the change of distributed capacitance with frequency and the drift of voltage divider resistor value with temperature are accurately characterized. By introducing the dung beetle optimization algorithm with compensation parameters as optimization variables for global optimization, and directly defining the voltage division ratio error as the optimization objective of the fitness function, the optimization capability and convergence speed of the algorithm are fully utilized. It can quickly lock the optimal frequency compensation parameters and temperature compensation parameters under complex wide frequency and variable temperature conditions, and finally effectively reduce the comprehensive influence of stray capacitance current fluctuations and temperature changes on the voltage division measurement results, significantly improving the measurement accuracy and stability of the high-voltage resistor voltage divider in the entire operating range.

[0016] Furthermore, the stray capacitance parameter error model of the present invention integrates the equivalent total distributed capacitance to ground, the distributed capacitance of the leads, the signal frequency, and the total resistance of the high and low voltage arms, comprehensively covering the coupling effect of internal and external distributed capacitance on the voltage division ratio, and providing an accurate physical basis for the quantification of frequency domain error.

[0017] Furthermore, the temperature parameter error model of the present invention introduces the temperature change coefficient and temperature change value of the high and low voltage arm resistors respectively, taking into account the difference in temperature drift characteristics of the resistors on both sides, so that the optimization result of the temperature compensation coefficient can more accurately match the temperature drift response of the actual voltage divider structure on both sides.

[0018] Furthermore, the fitness function of the present invention takes the maximum error across the entire frequency band as the optimization target, and incorporates frequency error and temperature error into a unified evaluation system to ensure that the worst deviation of the optimization result is minimized throughout the entire operating frequency band.

[0019] Furthermore, the boundary of the optimal foraging area in this invention adaptively shrinks as the dynamic reduction factor decreases linearly from 1 to 0, achieving a smooth transition from global wide-area exploration in the early stage of iteration to local fine development in the later stage, thereby improving the convergence accuracy and stability of the algorithm. Attached Figure Description

[0020] Figure 1 This is a flowchart of the multi-parameter optimization method for a high-voltage resistor divider according to an embodiment of the present invention; Figure 2 This is an equivalent circuit diagram of the high-voltage resistor divider according to an embodiment of the present invention; Figure 3 This is a flowchart of the multi-parameter optimization design algorithm for a high-voltage resistor divider according to an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0022] Example 1 This embodiment provides a multi-parameter optimization method for a high-voltage resistor divider, the process of which is as follows: Figure 1 As shown: Includes the following steps: S1. Based on the distributed capacitance in the high-voltage resistor divider, a stray capacitance parameter error model is constructed in the wide frequency domain; based on the relationship between the voltage divider resistance and temperature, an operating temperature parameter error model is constructed.

[0023] Specifically, error modeling is performed first. Based on the distributed capacitance in the high-voltage resistor divider, a stray capacitance parameter error model in the wide frequency domain is constructed. This stray capacitance parameter error model is determined based on the equivalent total distributed capacitance of the divider resistor to ground, the distributed capacitance of the high-voltage lead to the divider body, the signal frequency, and the total resistance of the high and low voltage arms.

[0024] Based on the relationship between the voltage divider resistance and temperature, an error model for the operating temperature parameter is constructed. Specifically, this error model is determined based on the temperature variation coefficients of the high-voltage arm resistor and the low-voltage arm resistor, their temperature variation values, and the nominal resistance values ​​of both the high- and low-voltage arms. Using these two models, the sources of error in the voltage divider are quantitatively characterized in both the frequency and temperature domains.

[0025] Specifically, the equivalent circuit of the high-voltage resistor divider is as follows: Figure 2 As shown, a high-voltage DC voltage is applied to the input terminal U1. The input current is divided into two paths: one path flows through the high-voltage arm resistor R1, and the other path flows into the ground through the equivalent total distributed capacitance C to ground; the current flowing through R1 continues to flow to the low-voltage arm resistor R2, and at this node it is split again, flowing to the output terminal U2 and leaking through the distributed capacitance C3 of the voltage divider body via the high-voltage terminal lead.

[0026] S2. Using the compensation parameters of the high-voltage resistor divider as optimization variables, initialize the search population of the dung beetle optimization algorithm, where each individual corresponds to a set of candidate values ​​for the compensation parameters.

[0027] After modeling is completed, the compensation parameters of the high-voltage resistor divider are used as optimization variables to initialize the search population for the dung beetle optimization algorithm, where each individual corresponds to a set of candidate values ​​for the compensation parameters. Further, the compensation parameters specifically include a frequency compensation coefficient and a temperature compensation coefficient. When initializing the dung beetle population, the location information of each individual dung beetle contains candidate values ​​in multiple dimensions, corresponding to the frequency compensation coefficient and the temperature compensation coefficient respectively, thus ensuring that each individual carries a complete set of compensation parameter combinations.

[0028] S3. Based on the stray capacitance parameter error model and the operating temperature parameter error model, define a fitness function with the voltage division ratio error as the optimization objective. Next, based on the stray capacitance parameter error model and the operating temperature parameter error model, a fitness function is defined with the voltage divider ratio error as the optimization objective. Further, the fitness function is determined by: calculating the error value of the high-voltage resistor divider at a given frequency and temperature; this error value is the absolute value of the sum of the outputs of the stray capacitance parameter error model and the temperature parameter error model, minus the absolute value of the compensation term calculated from the frequency compensation coefficient, temperature compensation coefficient, and the current frequency and temperature; and taking the maximum value among all frequency points within the entire preset operating frequency range as the fitness value of the dung beetle algorithm. This fitness design, targeting the maximum error across the entire frequency band, ensures that the worst deviation of the optimization result is minimized throughout the entire operating frequency band.

[0029] S4. The dung beetle optimization algorithm is used to iteratively optimize the search population and output the compensation parameters corresponding to the individual with the best fitness, which are used as the optimal compensation parameters for the high-voltage resistor divider.

[0030] The dung beetle optimization algorithm is used to iteratively optimize the search population, outputting the compensation parameters corresponding to the individual with the best fitness, which serve as the optimal compensation parameters for the high-voltage resistor divider. This iterative optimization process specifically includes: randomly dividing the initialized dung beetle population into four categories: rolling dung beetles, hatching dung beetles, foraging dung beetles, and thieving dung beetles; calculating the fitness of each category in each iteration and performing position updates according to their respective position update methods; repeating the calculation of fitness and position updates until the maximum number of iterations is reached, then outputting the compensation parameters corresponding to the individual with the best fitness.

[0031] During the aforementioned iterative optimization process, the position update methods for the four types of dung beetle individuals are as follows.

[0032] The position update of the dung beetle includes two scenarios: unobstructed and obstructed. For the unobstructed dung beetle, its next position is determined by its current position, previous position, worst-case global position, deflection coefficient, intensity coefficient, and a random number of -1 or 1. The worst-case global position is used to simulate distance measurements of light intensity changes. For the obstructed dung beetle, its next position is determined by its current position, previous position, and deflection angle. Position updates are not performed when the deflection angle is 0, π / 2, or π. This differentiated design allows the dung beetle to roll normally along an unobstructed path and adjust its direction of travel by dancing-like turns when encountering obstacles.

[0033] The position update method of the dung beetle is as follows: the position at the next moment is determined by the current position, two independent random vectors, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound, multiplied by each random vector, summed, and then added to the current position.

[0034] The position update method for foraging dung beetles is as follows: the position at the next moment is determined by the current position, two random vectors that follow a normal distribution, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound, multiplied by each normal random vector, summed, and then added to the current position.

[0035] The dung beetle's position is updated as follows: the position at the next moment is determined by the global best position, the distance from the current position to the global best position, a random vector that follows a normal distribution, and a preset constant. The constant is multiplied by the normal random vector and the distance and then added to the global best position.

[0036] Furthermore, the upper and lower bounds of the optimal foraging area are determined as follows: Calculations are performed based on the global best position, the dynamic reduction factor, and the preset upper and lower bounds of the search space. The lower bound of the optimal foraging area is the larger of the global best position multiplied by (1 minus the dynamic reduction factor) and the preset lower bound of the search space. The upper bound of the optimal foraging area is the smaller of the global best position multiplied by (1 plus the dynamic reduction factor) and the preset upper bound of the search space. The dynamic reduction factor decreases linearly from 1 to 0 with each iteration, thus ensuring that the foraging area is wider in the early stages of iteration to facilitate global exploration, and gradually shrinks as iterations progress to focus on a refined search near the global best position.

[0037] Accordingly, this embodiment also provides a multi-parameter optimization system for a high-voltage resistor divider. This system includes an error modeling module, a population initialization module, a fitness definition module, and an iterative optimization module. The error modeling module is used to construct a stray capacitance parameter error model in the wide frequency domain based on the distributed capacitance in the high-voltage resistor divider; and to construct an operating temperature parameter error model based on the relationship between the voltage divider resistance and temperature. The population initialization module is used to initialize the search population of the dung beetle optimization algorithm using the compensation parameters of the high-voltage resistor divider as optimization variables, where each individual corresponds to a set of candidate values ​​for the compensation parameters. The fitness definition module is used to define a fitness function with the voltage divider ratio error as the optimization objective based on the stray capacitance parameter error model and the operating temperature parameter error model. The iterative optimization module is used to iteratively optimize the search population using the dung beetle optimization algorithm, outputting the compensation parameters corresponding to the individual with the best fitness as the optimal compensation parameters for the high-voltage resistor divider.

[0038] Example 2 Based on Example 1, this example provides another method for multi-parameter optimization of a high-voltage resistor divider, including the following steps: T1. Construct a stray capacitance parameter error model and an operating temperature parameter error model for a high-voltage resistor divider in a wide frequency domain.

[0039] Optionally, step T1 specifically includes: T11, equivalent circuit of high-voltage resistor divider as follows Figure 2 As shown, considering the distributed capacitances C1 and C2 to ground of the voltage divider resistors in the high-voltage resistor divider, which cause uneven voltage distribution, and the distributed capacitance C3 of the high-voltage terminal lead to the divider body, the voltage division ratio under the condition of stray capacitance can be obtained as follows:

[0040] Where C=C g +C3, C g R is the equivalent total distributed capacitance of the voltage divider resistor to ground; R = R1 + R2, where R1 is the main voltage divider resistor and R2 is the sampling resistor; C3 is the distributed capacitance of the high-voltage lead to the voltage divider body; w is the signal frequency; U1 is the high-voltage input voltage and U2 is the low-voltage output voltage.

[0041] T12, the error model dominated by the change in distributed capacitance value is:

[0042] Where j is the imaginary unit, and jw is the purely imaginary part of the complex frequency.

[0043] T13, The error model mainly caused by the change in the value of the voltage divider resistor is:

[0044] Where, k HR k is the temperature variation coefficient of the high-voltage arm resistor. LR ΔT is the temperature variation coefficient of the low-voltage arm resistor; ΔT is the temperature variation value.

[0045] T2. Initialize the dung beetle population and calculate the fitness function of the dung beetle algorithm based on the total error of the voltage division ratio of the high-voltage resistor divider.

[0046] Optionally, step T2 specifically includes: T21. Initialize a population X of N dung beetles, where each individual represents a candidate solution, including frequency compensation coefficients kw1, kw2 and temperature compensation coefficients kt0, kt1, kt2; T22. Dung beetle populations were randomly divided into four categories: rolling dung beetle, hatching dung beetle, foraging dung beetle, and thieving dung beetle. T23. Calculate the error err of the high-voltage resistor divider at frequency w and temperature t, and take the maximum error over the entire frequency range as the fitness function of the dung beetle algorithm.

[0047] err=max{|errC(w)+errR(t)|-|kw1×w-kw2×w2-kw1×(kt0-t)-kw2×(kt1-t)2|} T3. Iterate through the dung beetle population until the maximum number of iterations is reached, then end the iteration to complete the multi-parameter optimization design of the high-voltage resistor divider.

[0048] Optionally, step T3 specifically includes: T31. Calculate the fitness of individual dung beetles; T32. Update the position of randomly generated unobstructed rolling dung beetles.

[0049]

[0050] Where t represents the current iteration number; X i (t) Let be the position information of the i-th dung beetle at the t-th iteration; k is the deflection coefficient, ranging from (0, 0.2]; b is the intensity coefficient, ranging from (0, 1); α is a random number taking the value -1 or 1; X W The worst-case position is represented globally; Δx is used to simulate changes in light intensity.

[0051] T33. Update the positions of the remaining obstacle-bound dung beetles.

[0052] Where θ is the deflection angle, ranging from [0, π]; the position is not updated when θ is 0, π / 2 or π.

[0053] T34. Update foraging area boundaries

[0054]

[0055] Among them, X b The optimal global position; Lb b This represents the lower bound of the optimal foraging area; Ub b R is the upper bound of the optimal foraging region; R = 1 - t / Tmax, where Tmax is the maximum number of iterations.

[0056] T35, Update the location of the dung beetle hatchling.

[0057] Here, b1 and b2 are two independent random vectors.

[0058] T36. Update the location of foraging dung beetles.

[0059] Where C1 and C2 are random vectors that follow a normal distribution.

[0060] T37, Updated location of the thief dung beetle.

[0061] Where g is a random vector that follows a normal distribution; S is a constant.

[0062] T38. Repeat T31 to T37. When the maximum number of iterations is reached, complete the dung beetle algorithm optimization to obtain the frequency compensation coefficient and temperature compensation coefficient, and complete the multi-parameter optimization design of the high-voltage resistor divider.

[0063] Furthermore, the multi-parameter optimization algorithm flow for the high-voltage resistor divider in this embodiment is as follows: Figure 3 As shown: Error models in the frequency and temperature domains are constructed. The dung beetle population is then initialized, and each set of compensation parameters is encoded as individual location information. Upon entering the iterative loop, the algorithm first calculates the fitness of each individual, then sequentially updates the position of the rolling dung beetle and dynamically adjusts the foraging area boundary, followed by updates to the positions of the hatching and foraging dung beetles. Finally, the algorithm determines whether to continue the iterative optimization or output the globally optimal compensation parameters and terminate the program by checking if the termination condition is met.

[0064] In summary, this invention provides a multi-parameter optimization method and system for a high-voltage resistor divider, which can suppress errors caused by frequency and temperature and improve the accuracy of voltage division measurement.

[0065] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0066] The order of the steps in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0067] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for multi-parameter optimization of a high-voltage resistive voltage divider, characterized in that, include: Based on the distributed capacitance in the high-voltage resistor divider, a stray capacitance parameter error model is constructed in the wide frequency domain; based on the relationship between the voltage divider resistance and temperature, an operating temperature parameter error model is constructed. Using the compensation parameters of the high-voltage resistor divider as optimization variables, the search population of the dung beetle optimization algorithm is initialized. The position information of each dung beetle in the search population corresponds to a set of candidate values ​​of the compensation parameters. Based on the stray capacitance parameter error model and the operating temperature parameter error model, a fitness function with the voltage division ratio error as the optimization objective is defined. The dung beetle optimization algorithm is used to iteratively optimize the search population and output the compensation parameters corresponding to the individual with the best fitness, which are then used as the optimal compensation parameters for the high-voltage resistor divider.

2. The high voltage resistive divider multi-parameter optimization method of claim 1, wherein, The compensation parameters specifically include the frequency compensation coefficient and the temperature compensation coefficient.

3. The high voltage resistive divider multi-parameter optimization method of claim 1, wherein, The stray capacitance parameter error model is determined based on the equivalent total distributed capacitance of the voltage divider resistor to ground, the distributed capacitance of the high-voltage end lead to the voltage divider body, the signal frequency, and the total resistance of the high and low voltage arms.

4. The multi-parameter optimization method for a high-voltage resistor divider according to claim 1, characterized in that, The operating temperature parameter error model is determined based on the temperature change coefficient of the high-voltage arm resistor, the temperature change coefficient and temperature change value of the low-voltage arm resistor, and the nominal resistance values ​​of the high and low voltage arms.

5. The high voltage resistive divider multi-parameter optimization method of claim 2, wherein, The fitness function is determined in the following ways: Calculate the error value of the high-voltage resistor divider at a given frequency and temperature; the error value is the absolute value of the sum of the output of the stray capacitance parameter error model and the output of the temperature parameter error model, minus the absolute value of the compensation term calculated by the frequency compensation coefficient, the temperature compensation coefficient and the current frequency and temperature; The maximum value among all error values ​​corresponding to all frequency points within the entire preset working frequency range is taken as the fitness value of the dung beetle algorithm.

6. The high voltage resistive divider multi-parameter optimization method of claim 1, wherein, The method of iteratively optimizing the search population using the dung beetle optimization algorithm includes: The initialized dung beetle population was randomly divided into four categories of individuals: rolling dung beetle, hatching dung beetle, foraging dung beetle, and thieving dung beetle. In each iteration, the fitness of each type of individual is calculated, and the position update is performed according to the corresponding position update method. Repeatedly calculate fitness and position updates until the preset maximum number of iterations is reached, then stop and output the compensation parameters corresponding to the individual with the best fitness.

7. The multi-parameter optimization method for a high-voltage resistor divider according to claim 6, characterized in that, The position update of the dung beetle includes position updates for unobstructed dung beetles and positions for dung beetles with obstacles. The next position of the unobstructed dung beetle is determined by the current position, the previous position, the worst-case position, the deflection coefficient, the intensity coefficient, and a random number with a value of -1 or 1. The worst-case position is used to simulate the distance measurement of light intensity changes. The next position of the dung beetle with obstacles is determined by the current position, the previous position, and the deflection angle. The position is not updated when the deflection angle is 0, π / 2, or π. The position update method of the dung beetle is as follows: the position at the next moment is determined by the current position, two independent random vectors, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound, multiplied by each random vector, summed, and then added to the current position. The position update method of the foraging dung beetle is as follows: the position at the next moment is determined by the current position, two random vectors that follow a normal distribution, the upper bound of the optimal foraging area, and the lower bound of the optimal foraging area. The current position is subtracted from the lower bound and the upper bound respectively, multiplied by each normal random vector, summed, and then added to the current position. The position update method of the dung beetle is as follows: the position at the next moment is determined by the global best position, the distance from the current position to the global best position, a random vector that follows a normal distribution, and a preset constant. The constant is multiplied by the normal random vector and the distance and then added to the global best position.

8. The high voltage resistive divider multi-parameter optimization method of claim 7, wherein, The upper and lower bounds of the optimal foraging area are determined in the following way: The calculation is performed based on the global optimal position, dynamic reduction factor, and preset upper and lower bounds of the search space. The lower bound of the optimal foraging area is the larger of the difference between the global best position multiplied by 1 and the dynamic reduction factor, and the preset lower bound of the search space. The upper bound of the optimal foraging area is the smaller of the sum of the global best position multiplied by 1 and the dynamic reduction factor, and the preset upper bound of the search space. The dynamic reduction factor decreases linearly from 1 to 0 with the number of iterations.

9. The high voltage resistive divider multi-parameter optimization method of claim 2, wherein, When initializing the dung beetle population, the location information of each individual dung beetle contains candidate values ​​in multiple dimensions, corresponding to the frequency compensation coefficient and the temperature compensation coefficient, respectively.

10. A high voltage resistive divider multi-parameter optimization system, comprising: include: The error modeling module is used to construct a stray capacitance parameter error model in a wide frequency domain based on the distributed capacitance in the high-voltage resistor divider; and to construct an operating temperature parameter error model based on the relationship between the voltage divider resistance and temperature. The population initialization module is used to initialize the search population of the dung beetle optimization algorithm with the compensation parameters of the high voltage resistor divider as optimization variables. The position information of each dung beetle in the search population corresponds to a set of candidate values ​​of the compensation parameters. The fitness definition module is used to define a fitness function with the voltage division ratio error as the optimization objective based on the stray capacitance parameter error model and the operating temperature parameter error model. The iterative optimization module is used to iteratively optimize the search population using the dung beetle optimization algorithm, and output the compensation parameters corresponding to the individual with the best fitness as the optimal compensation parameters of the high-voltage resistor divider.