Rapid calibration and tolerance analysis method for transmission extreme value frequency of high-frequency transformer

Abstract

The invention discloses a high-frequency transformer transmission extreme value frequency rapid calibration and tolerance analysis method, which comprises the following steps of collecting operation parameters of a high-frequency transformer in real time, performing standardized preprocessing, and integrating the preprocessed parameters to construct operation condition vectors in a unified format; defining a frequency response range of the high-frequency transformer, constructing a frequency search space based on the frequency response range, and establishing a target function for fitness evaluation based on the frequency search space; searching extreme value frequency points in the frequency search space by adopting a quantum evolutionary algorithm; based on the working condition vector and the extreme value frequency point, a working condition frequency deviation prediction model is constructed, and the working condition frequency deviation prediction model outputs an extreme value frequency shift offset; and calculating the dynamic tolerance bandwidth according to the predicted offset and the fluctuation range of the working condition vector. According to the invention, rapid and accurate calibration of the extreme value frequency of the high-frequency transformer and dynamic determination of the tolerance range can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of high-frequency transformer performance analysis and optimization, and in particular to a method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer. Background Art

[0002] In high-frequency power electronic equipment, high-frequency transformers are key components for energy transmission and electromagnetic isolation, and their transmission performance is highly dependent on the accurate matching of resonant frequency points. Traditional extreme frequency calibration methods usually rely on manual experience or fixed frequency point settings, and cannot fully consider the frequency offset phenomenon caused by changes in multiple factors such as voltage, current, temperature, and load conditions in the operating environment. In addition, existing frequency response testing methods mainly rely on traversal frequency sweeps, which are inefficient and costly, and cannot meet the system requirements of real-time adjustment and adaptive operation. Especially under complex working conditions, frequency drift cannot be ignored. If the extreme frequency is not accurately identified, it will significantly affect the efficiency and reliability of the system, and even cause resonant mismatch problems.

[0003] On the other hand, traditional methods for analyzing transformer frequency response often rely on empirically defined or fixed margins for tolerance bandwidth, failing to dynamically adjust the tolerance range based on current operating conditions. This results in conservative or insufficient settings, impacting system tuning and safety margins. In particular, lack of sophisticated modeling and prediction capabilities makes it difficult to properly quantify and dynamically adjust the tolerance under conditions of significant frequency offset, hindering the widespread application of high-frequency transformers in adaptive power supply systems. Summary of the Invention

[0004] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0005] To solve the above technical problems, the present invention provides the following technical solution: a method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer, comprising the following steps: Real-time collection of high-frequency transformer operating parameters, standardized preprocessing, and integration of preprocessed parameters into a unified format operating condition vector; Defining a frequency response range of the high-frequency transformer, constructing a frequency search space based on the frequency response range, and establishing an objective function for fitness evaluation based on the frequency search space; Using a quantum evolutionary algorithm to search for extreme frequency points in the frequency search space; Based on the operating condition vector and the extreme frequency point, a working condition frequency offset prediction model is constructed, and the working condition frequency offset prediction model outputs the extreme frequency shift offset; A dynamic tolerance bandwidth is calculated according to the predicted offset and the fluctuation range of the operating condition vector.

[0006] As a preferred solution of the method for rapid calibration and tolerance analysis of the high-frequency transformer transmission extreme frequency of the present invention, the operating condition vector is:

[0007] in, is the input voltage, is the input current, is the coil temperature, is the core temperature, In load state.

[0008] As a preferred solution of the method for rapid calibration and tolerance analysis of the high-frequency transformer transmission extreme frequency of the present invention, defining the frequency response range of the high-frequency transformer includes the following steps: Calculate the theoretical resonant frequency of a high-frequency transformer based on its equivalent topological parameters The initial frequency range is extended by ±20% with this frequency as the center. ; Based on the initial frequency range, a network analyzer is used to perform a frequency sweep test on the high-frequency transformer to obtain the reflection parameters. , and extract The frequency range is the final frequency response range .

[0009] As a preferred solution of the method for rapid calibration and tolerance analysis of the high-frequency transformer transmission extreme frequency of the present invention, wherein: obtaining the reflection parameter , and extract The specific steps of using the frequency interval as the final frequency response range are: Use a network analyzer to connect the high-frequency transformer to the test port. During the test, one end receives the excitation signal and the other end is connected to the reflection signal measurement channel. The step frequency is set within the initial frequency range to reflect the parameters of the high frequency transformer. Scan measurement, recording the reflection coefficient at each frequency point value; The reflection parameters in the test results Plot or store as frequency- Comparison table, for all frequency points The values ​​are traversed and the filter meets the All frequency point intervals; to satisfy The frequency band of the condition is the boundary, and the continuous frequency interval is extracted as the final frequency response range .

[0010] As a preferred solution of the method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer of the present invention, wherein: constructing a frequency search space based on the frequency response range includes the following steps: The frequency step interval Δf is set according to the actual frequency sweep capability of the network analyzer and the response characteristics of the transformer; The final frequency response range obtained by network analyzer screening , the entire frequency band is linearly divided with the set step interval Δf to obtain a set of frequency points arranged in ascending order ,in , , is the number of sampling frequencies; The frequency point set F is used as a basic search unit in the frequency search space to form a one-dimensional continuous search space.

[0011] As a preferred solution of the method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer according to the present invention, the specific steps of establishing an objective function for fitness evaluation based on the frequency search space are as follows: In the constructed frequency search space, for each frequency sampling point, two key indicator values ​​reflecting the transmission performance of the high-frequency transformer, namely, transmission gain and input return loss, are extracted as a basis for subsequent evaluation; Normalize or standardize the transmission gain value and reflection loss value of each frequency point so that the two parameters are in the same numerical range; Based on the normalized results of transmission gain and reflection loss, a comprehensive transmission characteristic discriminant function is designed as the objective function. The output value of the objective function is used as the fitness evaluation value of the current frequency point; The mathematical expression of the objective function is:

[0012] in: is the frequency point currently evaluated, obtained discretely in the frequency search space; The signal transmission gain from port 1 input to port 2 output is measured by a network analyzer at the frequency Measured at, it represents the transmission performance of the high-frequency transformer at this frequency; The reflection coefficient of port 1 input is measured by a network analyzer to represent the matching degree of the input end. The smaller the reflection, the better the matching. is the transmission gain; is the reflection coefficient; , It is an adjustable parameter, which can be set by the user or optimized through experience; is a very small positive number.

[0013] As a preferred solution of the method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer according to the present invention, the quantum evolutionary algorithm is a quantum annealing algorithm or a quantum genetic algorithm, and the search for the extreme frequency point includes the following steps: In the constructed frequency search space, each candidate frequency point is represented by quantum bit encoding to initialize the quantum individual population; Perform quantum measurement operations on each quantum individual in the initialized quantum individual population to collapse the quantum bit sequence of each quantum individual into the corresponding classical bit string, and decode and map the obtained classical bit string to obtain the corresponding frequency sampling points, and then calculate its fitness value in combination with the objective function; According to the size of the fitness value, excellent individuals are selected as the basis for population evolution; Based on the state difference between the current individual and the optimal individual, an adaptive quantum rotating gate is used to update and adjust the amplitude parameters of the quantum bit, guiding the probability distribution of the entire population to evolve towards the global optimal direction; Repeat the quantum measurement, fitness evaluation and quantum gate update operations for multiple rounds of iterative optimization. When the set maximum number of iterations is reached or the fitness convergence of the population meets the termination condition, the search is stopped and the frequency point with the largest fitness value is output as the calibrated extreme frequency point.

[0014] As a preferred solution of the method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer according to the present invention, the specific steps of using an adaptive quantum rotating gate to update and adjust the amplitude parameters of the quantum bit are as follows: For each quantum individual in the current population, extract the classical bit string obtained from its quantum measurement and compare it bit by bit with the classical bit string of the optimal individual with the highest fitness value in the current iteration. If the bit value on a certain bit is different, it is determined that the bit needs to be rotated and updated; For the bits that need to be rotated and updated, the rotation direction of the quantum rotation gate is set according to the difference between the current bit value and the target bit value, so that the quantum bit amplitude shifts towards the optimal solution direction; and the rotation angle is adaptively determined. .

[0015] For the quantum bits to be updated, based on the set rotation direction and rotation angle , perform amplitude update operation; Repeat the above rotation update operation to complete the amplitude parameter update of all quantum individuals in the current quantum population and obtain a new generation of quantum individual population.

[0016] As a preferred solution of the method for rapid calibration and tolerance analysis of the high-frequency transformer transmission extreme frequency of the present invention, the specific steps of constructing the operating frequency offset prediction model are: Under multiple operating conditions, obtain the corresponding operating condition vector And the extreme frequency calibrated under this working condition , with the reference frequency under standard working conditions As a reference, calculate the frequency offset , forming a sample pair ; All the formed sample pairs are combined into a sample set, which is then randomly divided into a training set and a validation set. The training set is used for model construction and parameter fitting, and the validation set is used to evaluate the model prediction accuracy. Standardize the parameters in the operating condition vector and use principal component analysis or correlation analysis to reduce the dimension of variables; A polynomial regression model is constructed to express the nonlinear mapping relationship between the frequency offset and the operating condition vector. The model parameters are solved using the least squares method, and the prediction accuracy is evaluated on the validation set.

[0017] As a preferred solution of the method for rapid calibration and tolerance analysis of the high-frequency transformer transmission extreme frequency of the present invention, the calculation formula of the dynamic tolerance bandwidth is:

[0018] in: 、 The value of is positively correlated with the fluctuation amplitude of the working condition. The downward tolerance indicates the maximum allowable frequency deviation of the transformer under the current working condition. The upward tolerance indicates the maximum allowable frequency deviation value for upward deviation; is the corrected extreme frequency, .

[0019] Beneficial effects of the present invention: 1. The present invention constructs an operating condition vector in a unified format and uses a quantum evolutionary algorithm to search for extreme frequency points in the constructed frequency search space, significantly improving the search efficiency and accuracy of extreme frequencies and overcoming the problems of low efficiency and poor accuracy of traditional methods based on traversal or manual experience.

[0020] 2. The present invention constructs an operating condition frequency offset prediction model, introduces a mapping mechanism between operating condition vectors and frequency offsets, and combines sample set training with variable dimensionality reduction, so that the system can automatically correct extreme frequency, has strong generalization ability, and significantly improves adaptability and stability under different operating conditions.

[0021] 3. The present invention further calculates the dynamic tolerance bandwidth based on frequency calibration. The present invention further calculates the dynamic tolerance bandwidth based on frequency calibration. ,in Positively correlated with the amplitude of operating condition fluctuations, it has practical physical significance, making the calibrated frequency more robust and engineering feasible. The bandwidth clearly covers the possible range of frequency drift, providing a quantitative basis for subsequent filter design or system matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them: Figure 1 This is a flowchart of a method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer of the present invention.

[0023] Figure 2 The present invention provides a flowchart for searching for extreme frequency points in a method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer. DETAILED DESCRIPTION

[0024] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0026] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0027] Furthermore, the present invention is described in detail with reference to schematic diagrams. For ease of illustration, when describing the embodiments of the present invention, cross-sectional views illustrating device structures may be partially enlarged and not to scale. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of protection of the present invention. Furthermore, in actual production, the three-dimensional dimensions of length, width, and depth should be included.

[0028] Example 1 Reference Figure 1-2 , as one embodiment of the present invention, provides a method for rapid calibration and tolerance analysis of the transmission extreme frequency of a high-frequency transformer, comprising the following steps: S1: Real-time collection of high-frequency transformer operating parameters, standardized preprocessing, and integration of the preprocessed parameters into a unified format operating condition vector.

[0029] It should be noted that operating parameters include but are not limited to input voltage, input current, coil temperature, core temperature, and load status. These can be acquired in real time through voltage sampling circuits, current transformers, and temperature sensors. The load status can be calculated based on the real-time collected voltage and current data to obtain load power or impedance. These operating parameters can comprehensively characterize the actual operating conditions of the equipment from the input end, the transformer body, to the load side. Among them: input voltage and current are used to reflect the power supply conditions and load level; coil and core temperatures characterize the impact of thermal changes on the electromagnetic characteristics of the device; and load status is used to supplement the effect of the output side boundary on frequency drift. These parameters are highly correlated with the extreme frequency offset, ensuring that the subsequent operating condition frequency offset prediction model has good expressiveness and generalization capabilities.

[0030] Preprocessing includes normalization and noise filtering to eliminate the dimensional differences between different physical quantities. At the same time, measurement noise is suppressed through low-pass filters or wavelet noise reduction algorithms to improve the accuracy of working condition vector construction.

[0031] Specifically, the operating condition vector is:

[0032] in, is the working condition vector, is the input voltage, is the input current, is the coil temperature, is the core temperature, In load state.

[0033] It should be noted that the above step S1 not only realizes the comprehensive characterization of the input side, electromagnetic body and output load state of the high-frequency transformer through real-time collection and standardized preprocessing of the high-frequency transformer operating parameters and the construction of a unified format operating condition vector, but also effectively eliminates the dimensional interference and measurement noise influence between different physical quantities, thereby providing a high-quality and generalizable feature input basis for the subsequent modeling of the operating condition frequency offset prediction model, significantly improving the prediction accuracy and adaptability of the model.

[0034] S2: Define the frequency response range of the high-frequency transformer, construct a frequency search space based on the frequency response range, and establish an objective function for fitness evaluation based on the frequency search space.

[0035] Specifically, defining the frequency response range of a high-frequency transformer includes the following steps: Calculate the theoretical resonant frequency of a high-frequency transformer based on its equivalent topological parameters The initial frequency range is extended by ±20% with this frequency as the center. ; Based on the initial frequency range, a network analyzer is used to perform a frequency sweep test on the high-frequency transformer to obtain the reflection parameters. , and extract The frequency interval is taken as the final frequency response range, which provides an effective search space for subsequent extreme frequency calibration.

[0036] It should be noted that the equivalent topology parameters refer to the key circuit elements used to describe the internal electrical structure characteristics of the high-frequency transformer, including but not limited to the primary / secondary inductance, capacitance and load coupling coefficient. These parameters can be obtained from the equivalent circuit modeling of the device and are the main determinants affecting its resonant behavior. By selecting these parameters as input to calculate the theoretical resonant frequency It can not only accurately reflect the resonant position of the transformer under ideal conditions, but also provide a scientific initial range for subsequent frequency response tests, greatly reducing the frequency search blind area, and improving the pertinence of frequency distribution and the efficiency of frequency calibration.

[0037] In a specific embodiment, the theoretical resonant frequency The calculation can be modeled based on the topological parameters of the high-frequency transformer. Taking the equivalent series resonant structure as an example, the theoretical resonant frequency It can be calculated as follows:

[0038] Among them, L represents the equivalent inductance and C represents the equivalent capacitance, and their values ​​can be extracted through preliminary testing. For more complex topologies, such as push-pull and full-bridge, equivalent T or The resonant frequency point is extracted by using a transfer function method, which is a well-known method for those skilled in the art and will not be described in detail here.

[0039] It should be noted that the theoretical resonant frequency is calculated by the above method The main function is to quickly and accurately estimate the theoretical resonant frequency of the high-frequency transformer under ideal electrical conditions based on the equivalent topological parameters of the high-frequency transformer without actual measurement. , and then expand the initial frequency range by ±20% with this frequency as the center, significantly improving the pertinence and efficiency of subsequent network analyzer sweep frequency tests, avoiding the waste of test resources and noise interference caused by blind scanning of the entire frequency band, and also providing a more focused search space and a highly reliable data basis for the subsequent rapid search and precise calibration of extreme frequency points.

[0040] In a specific embodiment, the reflection parameter is obtained , and extract The specific steps of using the frequency interval as the final frequency response range are: Use a network analyzer to connect the high-frequency transformer to the test port. During the test, one end receives the excitation signal and the other end is connected to the reflection signal measurement channel. In the above initial frequency range The internal setting step frequency (such as 1kHz) is used to adjust the reflection parameters of the high frequency transformer. Scan measurement, recording the reflection coefficient at each frequency point value; The reflection parameters in the test results (in dB) Plot or store as frequency- Comparison table, for all frequency points The values ​​are traversed and the filter meets the All frequency point intervals; to satisfy The frequency band of the condition is the boundary, and the continuous frequency interval is extracted as the final frequency response range; if there are multiple discontinuous frequency bands that meet the conditions, the frequency band containing the reflection parameters can be selected The maximum continuous satisfaction of the minimum value corresponding to the frequency point Frequency response range of the conditions, ensuring that the range covers the optimal transmission performance point; The final frequency response range is recorded as , as the effective input frequency domain for subsequent extreme frequency search and objective function definition.

[0041] It should be noted that in the above-mentioned final frequency response range acquisition step, a network analyzer is used to perform point-by-point frequency scanning on the high-frequency transformer, and the frequency band with reflection loss less than -10dB is screened out as the effective frequency response range. This avoids the test redundancy and noise interference caused by blind scanning of the entire frequency band, ensuring that the selected frequency range truly covers the area with optimal transmission performance of the transformer, thereby improving the positioning accuracy and calculation efficiency of the subsequent extreme frequency search, and laying a solid foundation for the accurate construction of the objective function and the accurate calibration of the extreme frequency points.

[0042] Specifically, constructing a frequency search space based on the frequency response range includes the following steps: According to the actual frequency sweep capability of the network analyzer and the transformer response characteristics, set an appropriate frequency step interval Δf (such as 1 kHz, 500 Hz, or less) to ensure that the frequency sampling points are dense enough to cover the key details and local extreme value change trends in the frequency response curve, and avoid missing extreme points due to insufficient resolution; The final frequency response range obtained by network analyzer screening , the entire frequency band is linearly divided with the set step interval Δf to obtain a set of frequency points arranged in ascending order ,in , , is the number of sampling frequencies; The above frequency point set F is used as a basic search unit in the frequency search space to form a one-dimensional continuous search space.

[0043] It should be noted that the transformer response characteristics refer to the behavioral characteristics of parameters such as transmission gain and reflection loss of a high-frequency transformer in the frequency domain as they change with frequency. They are usually manifested as sharp resonance peaks or valleys (i.e., resonance points) near a certain frequency, and the response curve fluctuates violently and changes at a high rate in this area.

[0044] To effectively capture these resonant behaviors, especially ensuring frequency sampling accuracy in steep response regions, the frequency step interval must be set based on the response characteristics. Generally speaking, for response curves with narrow bandwidths or rapid frequency changes, a smaller frequency step interval (e.g., ≤500 Hz) should be used to avoid missing extreme points. For broadband transformers with smoother response changes, the step interval can be appropriately relaxed (e.g., 1-5 kHz) to balance sampling efficiency and accuracy.

[0045] In summary, by setting an appropriate frequency step interval based on the actual frequency sweep capability of the network analyzer and the response characteristics of the high-frequency transformer, and constructing a frequency search space by linearly dividing the final frequency response range, it is possible to ensure that the frequency sampling points have a sufficient density in the area where the transmission response changes sharply, effectively capture the changing trends of local extreme values ​​such as the resonance peak, and avoid missing key frequency points due to excessively large steps, providing accurate and reliable basic support for the subsequent construction of the objective function.

[0046] In a specific embodiment, the specific steps of establishing the objective function for fitness evaluation based on the frequency search space are: In the constructed frequency search space, the transmission gain and input return loss corresponding to each frequency sampling point are extracted as the key indicators reflecting the transmission performance of the high-frequency transformer, which serve as the basis for subsequent evaluation. Normalize or standardize the transmission gain value and reflection loss value of each frequency point so that the two parameters are in the same numerical range; Based on the normalized results of transmission gain and reflection loss, a comprehensive transmission characteristic discriminant function is designed as the objective function. The output value of the above objective function is used as the fitness evaluation value of the current frequency point to measure the quality of the comprehensive transmission performance of the frequency point. The larger the fitness value, the closer the point is to the ideal transmission characteristics, that is, high transmission gain and low reflection loss.

[0047] It should be noted that the objective function is used to comprehensively evaluate the transmission performance of the high-frequency transformer at each frequency point, and its output value is used as a fitness indicator for the quantum evolutionary algorithm described in step S3 to drive it to search for the optimal extreme frequency point in the frequency search space.

[0048] In a specific embodiment, the mathematical expression of the objective function is:

[0049] in: is the frequency point of the current evaluation; The signal transmission gain from port 1 input to port 2 output is measured by a network analyzer at the frequency Measured at, it represents the transmission performance of the high-frequency transformer at this frequency; The input reflection coefficient of port 1 is also measured by a network analyzer to represent the matching degree of the input end. The smaller the reflection, the better the matching. is the transmission gain; is the reflection coefficient; , It is an adjustable parameter, which can be set by the user or optimized through experience; is a very small positive number; This function comprehensively measures the transmission efficiency and the degree of matching of the input end. The larger the value, the better the transmission performance at the frequency point. Experimental results show that when =1, , the function is stable and effective in identifying extreme frequency points, and is suitable for the subsequent frequency calibration and tolerance analysis tasks of this application.

[0050] The above steps establish an objective function for fitness evaluation based on the frequency search space, which can establish a unified numerical evaluation criterion between the transmission performance indicators of the high-frequency transformer (such as transmission gain and input reflection loss), realize objective quantitative judgment of the transmission characteristics of each frequency point, and make the quality of the frequency point clearly expressed in numerical form; in particular, the dimensional interference between different physical quantities is avoided through standardization processing, and the adjustment weight and the minimum positive factor are introduced into the objective function, so that the objective function has good discrimination and sensitivity while maintaining numerical stability, which can effectively reflect the relative quality of the frequency point in the comprehensive transmission performance, and thus provide a discriminative fitness evaluation basis for the subsequent quantum evolutionary algorithm, thereby improving the accuracy and computational efficiency of the extreme frequency point search, and ensuring the stability and repeatability of the entire calibration process.

[0051] In summary, in step S2, the theoretical resonant frequency is calculated based on the equivalent topological structure parameters of the high-frequency transformer, and the final frequency response range is screened out in combination with the measured reflection parameters of the network analyzer. Then, an appropriate frequency step interval is set according to the response characteristics of the transformer to construct a frequency search space. Finally, an objective function that integrates the relationship between the transmission gain and the reflection loss weight is designed. This can not only significantly improve the efficiency and accuracy of subsequent frequency extreme point searches, but also effectively avoid computational redundancy and error diffusion caused by blind scanning of the entire frequency band.

[0052] S3: using a quantum evolutionary algorithm to search for extreme frequency points in the frequency search space; Specifically, the quantum evolutionary algorithm is a quantum annealing algorithm or a quantum genetic algorithm, which searches for extreme frequency points, including the following steps: In the constructed frequency search space, each candidate frequency point is represented by quantum bit encoding to initialize the quantum individual population; Perform quantum measurement operations on each quantum individual in the initialized quantum individual population to collapse the quantum bit sequence of each quantum individual into the corresponding classical bit string, and decode and map the obtained classical bit string to obtain the corresponding frequency sampling points, and then calculate its fitness value in combination with the objective function; According to the size of the fitness value, excellent individuals are selected as the basis of population evolution. The higher the fitness value, the closer the frequency point is to the extreme point with the best transmission characteristics. Based on the state difference between the current individual and the optimal individual, an adaptive quantum rotating gate is used to update and adjust the amplitude parameters of the quantum bit, guiding the probability distribution of the entire population to evolve towards the global optimal direction, thereby enhancing the global optimization ability and avoiding falling into the local optimal direction; Repeat the quantum measurement, fitness evaluation and quantum gate update operations for multiple rounds of iterative optimization. When the set maximum number of iterations is reached or the fitness convergence of the population meets the termination condition, the search is stopped and the frequency point with the largest fitness value is output as the calibrated extreme frequency point.

[0053] In a specific embodiment, each candidate frequency point is represented by quantum bit encoding, and the specific steps of initializing the quantum individual population are: According to the upper and lower limits of the constructed frequency search space and frequency step interval, calculate the number of frequency sampling points N, and determine the minimum quantum bit code length required to uniquely identify all frequency points; According to the determined minimum quantum bit encoding length, several quantum individuals are initialized. Each quantum individual consists of m quantum bits in superposition state, where the initial amplitude parameter of each quantum bit is , to ensure equal probability coverage of the entire search space in the initial stage; Arrange a number of quantum individuals in sequence according to the coding length to construct a two-dimensional matrix structure.

[0054] It should be noted that the calculation formula for the number of frequency sampling points N is: ,in represents the interval between two adjacent frequency sampling points, and N represents the total number of sampling points to be covered within the frequency search range. To ensure that each frequency point can be uniquely identified and facilitate encoding and decoding through the quantum bit sequence, all frequency sampling points must be numbered in binary format. To this end, the minimum quantum bit encoding length L should satisfy the following relationship: , that is, the encoding length L satisfies The minimum integer to ensure that the coding space covers all candidate frequency points. For example, when N is 10, , that is, the minimum quantum bit encoding length L is 4, then , which means that at least 4 binary bits (4 quantum bits) are required to uniquely mark up to 16 frequency points, ensuring that each frequency point has an independent code without conflict or duplication.

[0055] It should be noted that each quantum individual is composed of m quantum bits in superposition state, where m is the quantum bit encoding length L required to uniquely identify all frequency points. The reason for this setting is that in the quantum evolution algorithm, each individual needs to uniquely identify a candidate solution, and the "candidate solution" in this application is a frequency sampling point. Since there are N frequency points in the frequency search space, in order to achieve unique encoding of each frequency point, no less than Therefore, if each frequency point is represented by a quantum bit string, the number of quantum bits m contained in the string must satisfy ,Right now If m < L, some frequency points will not be covered by the coding and the search space will be incomplete; if m > L, redundant bit width will be introduced, affecting efficiency.

[0056] It should be noted that the quantum state of each quantum bit can be expressed as: ,in

[0057] Substitute the initial amplitude parameters of the quantum bit into have to:

[0058] This means that each quantum bit has the ability to collapse with equal probability in the 0 and 1 states, so that the coding combinations of all frequency points are likely to appear in the initial stage, achieving a global uniform exploration of the search space. If the initial amplitude parameter is biased towards a certain state, such as >> , which will cause the search space to be skewed and easily fall into the local optimal solution. = It can prevent the search from falling into local optimality and provide better initial conditions for subsequent iterative optimization.

[0059] It should be noted that the specific steps for constructing a two-dimensional matrix structure are as follows: set the quantum population size to P, that is, the total number of initialized quantum individuals is P, and according to the set quantum population size P and the minimum quantum bit encoding length L, construct a two-dimensional matrix structure with a dimension of P×L, where each row represents a quantum individual, and each column represents a single quantum bit at the corresponding position of the individual. The matrix elements can also be marked with the initial amplitude state of the quantum bit. , defaulting to a superposition state of equal amplitude and phase, thus completing the structured expression of the initialized quantum individual population. This two-dimensional matrix structure serves as the basis for subsequent quantum measurement, fitness evaluation, and revolving gate operations, ensuring that the quantum evolution process has unified and efficient data support.

[0060] In the process of initializing the quantum individual population, by using quantum bit encoding to represent each candidate frequency point in the constructed frequency search space and initializing a quantum individual population composed of several quantum bits in a superposition state, it can not only ensure full coverage and accurate identification of the entire search space in the frequency point encoding stage, but also help to form a population structure with globally uniform distribution characteristics in the initial stage of the quantum evolutionary algorithm, thereby improving search efficiency, enhancing global optimization capabilities, and effectively avoiding the algorithm from falling into local optimal solutions.

[0061] In a specific embodiment, the specific steps of performing quantum measurement and fitness calculation on each quantum individual in the initialized quantum individual population are as follows: First, a quantum measurement operation is performed on the quantum bit string composed of m quantum bits in superposition state in each quantum individual. Through the quantum measurement operation, each quantum bit is changed from superposition state to Random collapse to the classical state or , where the collapse result follows the amplitude probability distribution, that is: ,

[0062] After the measurement is completed, each quantum individual will be represented by a certain 0-1 sequence, forming the corresponding classical bit string.

[0063] Then, perform binary decoding on each measured classical bit string to obtain its index number in the frequency search space. Assume that a classical bit string is , then the corresponding decimal index value is calculated as follows:

[0064] in, k Indicates the corresponding frequency sampling point index value; L represents the minimum quantum bit encoding length; i Indicates the bit number in the classic bit string, with a value range of 0≤ i ≤ L -1, low position first, high position last; Indicates the current i The corresponding binary weight; After quantum measurement i The value of a quantum bit (0 or 1).

[0065] The index value k Corresponding to the first k Frequency sampling points, convert the index value to the corresponding frequency value: ,in, Indicates the frequency step interval, is the lower bound of the search space.

[0066] Finally, the obtained frequency sampling points Substitute it into the objective function pre-constructed in step S2 and calculate its fitness value. The output value of the objective function is the fitness value of the frequency point represented by the quantum individual. The higher the fitness, the better the transmission performance of the frequency point.

[0067] Exemplary: Assuming minimum qubit encoding length L =4, that is, each frequency point is encoded by 4 quantum bits. After a quantum individual is quantum measured, the following classical bit string is obtained:

[0068] Calculated according to the index value formula:

[0069] Therefore, the frequency sampling point index value corresponding to this quantum individual is 11, and the index value is converted to the frequency value: , assuming , =0.9, =0.3, =1, ; After substituting into the objective function:

[0070] therefore, =0.3333, which is the frequency sampling point =210MHz fitness value.

[0071] In summary, by measuring and decoding quantum individuals in superposition states and combining them with objective functions for fitness calculation, not only can the quantum bit state be mapped to specific frequency points, but its transmission performance can also be quantitatively evaluated, making it easier to distinguish and sort the pros and cons of each candidate frequency point, thereby providing a scientific basis for subsequent individual screening and quantum evolution, and improving the global optimization capability and efficiency of the search process.

[0072] In a specific embodiment, the specific steps of using an adaptive quantum rotating gate to update and adjust the amplitude parameters of a quantum bit are as follows: For each quantum individual in the current population, the classical bit string obtained from its quantum measurement is extracted and compared bit by bit with the classical bit string of the optimal individual with the highest fitness value in the current iteration. If the bit value in a certain position is different, it is determined that the position needs to be rotated and updated.

[0073] For the bits that need to be rotated and updated, the rotation direction of the quantum rotation gate (clockwise or counterclockwise) is set according to the difference between the current bit value and the target bit value, so that the quantum bit amplitude shifts towards the optimal solution direction; and the rotation angle is adaptively determined. .

[0074] For the quantum bits to be updated, based on the set rotation direction and rotation angle , perform the following amplitude update operation:

[0075] By adjusting the amplitude parameters, this quantum bit can collapse to the target bit value (0 or 1) with a higher probability in the next quantum measurement; for example: the current individual's bit string is 1010; the optimal individual's bit string is 1110; then the first bit (the second one from the left) has the value: currently 0, optimally 1; inconsistent, requires rotation and update.

[0076] Repeat the above rotation update operation to complete the amplitude parameter update of all quantum individuals in the current quantum population and obtain a new generation of quantum individual population.

[0077] It should be noted that the bit value refers to the classical binary value that each quantum bit collapses into after measurement, that is, 0 or 1. The adaptive reduction is performed according to the preset decreasing function according to the iteration rounds, so that the search disturbance ability is larger in the early stage of the algorithm to promote global exploration, and gradually decreases with the increase of the number of iterations to enhance the local convergence accuracy.

[0078] It should also be noted that the method for determining the rotation direction of the revolving door is: if the bit values ​​are inconsistent, the rotation direction is determined according to the difference direction, wherein, when the current bit value is 0 and the optimal bit value is 1, counterclockwise rotation is adopted to enhance State amplitude; otherwise, rotate clockwise to enhance State amplitude.

[0079] For example, assume: The current bit string of the quantum individual is 1010; The bit string of the current best individual in the population is 1110; We take the first bit of the 4-bit string as an example to illustrate the rotation update process; The initial amplitude of the first quantum bit is: , and satisfies .

[0080] Then, the current individual 1010 and the optimal individual 1110 are compared bit by bit. The first bit (the second from the left) is: The current bit value is 0, and the optimal bit value is 1 → they are different and need to be rotated and updated.

[0081] The current bit value is 0, and the target bit value is 1 → indicating that the probability of the quantum state needs to be Then, set the rotation direction to counterclockwise (let Increase, decrease).

[0082] Adaptive setting of rotation angle , such as using larger values ​​in the early iterations (such as ), which gradually decreases to .

[0083] Then, perform the revolving door update operation as follows: The original amplitude vector of the quantum bit is:

[0084] Calculate the rotation result:

[0085] The updated amplitude status is:

[0086] At this time, the bit is more likely to collapse to 1 in subsequent measurements, that is, to move closer to the optimal individual.

[0087] The above judgment and update are performed on all bits of all quantum individuals in the population in turn to form a new generation of quantum population. The iterative optimization process is repeated until the set termination condition is reached.

[0088] In summary, the use of adaptive quantum rotating gates to update and adjust the amplitude parameters of quantum bits can guide the quantum bit state to converge towards the current optimal solution, while ensuring the overall exploration capability of the population and dynamically adjusting the search perturbation amplitude, so that quantum individuals gradually approach the global optimal solution in multiple rounds of iterations, thereby effectively improving the convergence efficiency of frequency point search and the ability to obtain the global optimal solution, avoiding falling into local optimal solutions and enhancing the ability to accurately calibrate the extreme points of transmission characteristics.

[0089] In a specific embodiment, the maximum number of iterations may be determined according to one of the following methods: The empirical setting method is to set a fixed maximum number of iterations based on the actual problem scale (such as the frequency search space size N, the quantum bit encoding length L) and the computing power of the device; Adaptive adjustment method, that is, dynamically record the change range of the optimal fitness value in each round during the operation of the quantum evolutionary algorithm. When the fitness does not improve significantly after several consecutive rounds (such as 5 or 10 rounds), it will automatically trigger the stop in advance; The search space correlation setting, that is, the maximum number of iterations can be set to several times the number of search space points N, to ensure that all frequency points have sufficient exploration opportunities in a probabilistic sense.

[0090] In a specific embodiment, the fitness convergence termination condition is determined as follows: a threshold is set, and if the change range of the optimal fitness in several consecutive iterations is less than or equal to the threshold, it is considered to be converged.

[0091] In summary, step S3 is based on the quantum evolution algorithm. By introducing quantum bit encoding and initialization population strategy in the frequency search space, full coverage and unique identification of candidate frequency points are achieved. Through quantum measurement mapping, fitness evaluation, adaptive revolving door update and multiple rounds of iterative optimization process, the quantum population is dynamically guided to converge towards the optimal frequency point. This not only enhances the global search capability and avoids falling into local optimality, but also improves the accuracy and optimization efficiency of frequency point extreme value identification, significantly improving the global optimization capability and stability of the frequency calibration process in complex search space.

[0092] S4: Based on the operating condition vector and the extreme frequency point, a working condition frequency offset prediction model is constructed, and the working condition frequency offset prediction model outputs the extreme frequency shift offset; Specifically, the specific steps for constructing the operating frequency offset prediction model are as follows: Under multiple operating conditions, obtain the corresponding operating condition vector And the extreme frequency calibrated under this working condition , with the reference frequency under standard working conditions As a reference, calculate the frequency offset , forming a sample pair ; All the formed sample pairs are combined into a sample set, which is then randomly divided into a training set and a validation set. The training set is used for model construction and parameter fitting, and the validation set is used to evaluate the model prediction accuracy. Standardize the parameters in the operating condition vector and use principal component analysis or correlation analysis to reduce the dimension of variables; A polynomial regression model is constructed to express the nonlinear mapping relationship between the frequency offset and the operating condition vector. The model parameters are solved using the least squares method, and the prediction accuracy is evaluated on the validation set.

[0093] In a specific embodiment, the expression of the polynomial regression model is:

[0094] Used to dynamically predict the frequency offset of high-frequency transformers under target operating conditions , thereby improving the accuracy of extreme frequency calibration.

[0095] in, is the constant term coefficient, obtained by fitting historical sample data; is the first-order linear coefficient, indicating the working condition vector The linear relationship between dimension and frequency offset is obtained by fitting the training data; is the second-order cross-term coefficient, indicating the Hedi The effect of the product between the components of each working condition on the frequency offset is fitted by historical samples; the model parameters are obtained by training sample data; Indicates the working condition parameters (normalized); Indicates the working condition parameters (normalized); It is the cross term of the operating parameters, reflecting the joint effect between the two operating parameters.

[0096] During the operation of a high-frequency transformer, its transmission extreme frequency will shift due to the influence of operating conditions such as temperature, voltage, current and load changes. Therefore, a model that can dynamically predict the extreme frequency offset based on the operating conditions is needed to achieve the extreme frequency Dynamic correction improves the adaptability and accuracy of frequency calibration and provides quantitative offset for tolerance analysis basic support.

[0097] For example: a certain model of high-frequency transformer is selected, and tests are performed under different input conditions to collect operating parameters such as input voltage, input current, coil temperature, core temperature, and load resistance.

[0098] All parameters are normalized to the minimum and maximum values ​​to form a five-dimensional operating condition vector ; like 、 、 、 、 、 、 ; For simplicity, only some cross terms are retained (as follows): 、 ; Substitute into the working condition frequency offset prediction model for calculation: Sum of linear terms: + + + + =0.10×0.74+0.08×0.62+(-0.06)×0.55+0.12×0.48+0.04×0.70=0.1762 Sum the cross terms (only two are kept in this example): + =0.05×0.74×0.62+(-0.03)×0.05×0.48=0.01502 Add them all up:

[0099] above Only two cross terms are retained in the summation, and the actual result is =0.37MHz; Therefore, in practical applications, after embedding the model, the user only needs to input the current working condition vector , the model automatically outputs the .

[0100] In summary, step S4 constructs a working condition frequency offset prediction model, which can dynamically predict the offset of the transmission extreme frequency based on the current operating condition of the high-frequency transformer, thereby realizing adaptive correction of the calibration frequency, improving the accuracy and adaptability of frequency calibration under different working conditions, and providing a quantitative offset basis for subsequent tolerance analysis, thereby enhancing the stability and reliability of the overall method in the actual operating environment.

[0101] S5: Calculating a dynamic tolerance bandwidth according to the predicted offset and the fluctuation range of the operating condition vector.

[0102] In a specific embodiment, the calculation formula of the dynamic tolerance bandwidth is:

[0103] in: 、 The value of is positively correlated with the fluctuation amplitude of the working condition. The downward tolerance indicates the maximum allowable frequency deviation of the transformer under the current working condition. The upward tolerance indicates the maximum allowable frequency deviation value for upward deviation; is the corrected extreme frequency, .

[0104] By dynamically calculating the frequency tolerance bandwidth, the frequency calibration results can be made more adaptable and robust under different operating conditions, effectively improving the system's frequency control and error compensation capabilities and enhancing the transmission reliability during transformer operation.

[0105] In a specific embodiment, 、 The steps for quantifying the value of are as follows: for each parameter that constitutes the operating condition vector, such as temperature, current, voltage, load, etc., within a pre-set time window, the standard deviation or maximum deviation is used to measure the fluctuation range of each parameter. Taking into account the importance of each operating condition parameter, by assigning corresponding weights to different components, the weighted square sum calculation method is used to obtain the overall operating condition fluctuation amplitude index. Among them, the weight of each component can be determined based on historical experience or sensitivity analysis. Finally, based on the calculated overall operating condition fluctuation amplitude, the linear mapping method is used to determine the δ value. The linear mapping coefficient used here is an empirical fitting coefficient, which reflects the safety margin in the offset direction and can be obtained through sample data training or engineering practice calibration.

[0106] For example, under a typical operating condition: The extreme frequency points calibrated by the quantum evolutionary algorithm are: =210.0kHz.

[0107] The frequency offset calculated by the frequency offset prediction model based on the current operating condition vector is: =-1.8kMz.

[0108] The corrected extreme frequency is:

[0109] At the same time, considering the certain volatility of the current operating conditions, it is set as follows: (corresponding to the tolerance band for downward frequency shift) (corresponding to the tolerance band for upward frequency deviation) The final dynamic tolerance bandwidth is:

[0110] The bandwidth It is used to cover the possible range of changes in extreme frequencies under the current operating conditions, ensuring sufficient robustness and fault tolerance when performing subsequent filter design, impedance matching, or transmission performance evaluation based on this frequency bandwidth, and effectively preventing system performance degradation or communication failure due to frequency offset.

[0111] In summary, in step S5, the offset output by the operating condition frequency offset prediction model and the fluctuation range of the operating conditions are introduced to dynamically calculate the tolerance bandwidth of the calibration frequency. This can not only effectively reflect the frequency offset trend caused by fluctuations in operating conditions such as temperature, current, voltage, and load during actual operation of the high-frequency transformer, but also significantly improve the accuracy and reliability of the frequency calibration results under different operating conditions compared to fixed tolerance settings.

[0112] Therefore, the present invention significantly improves the search efficiency and accuracy of extreme frequencies by constructing an operating condition vector in a unified format and using a quantum evolutionary algorithm to search for extreme frequency points in the constructed frequency search space, overcoming the problems of low efficiency and poor accuracy of traditional methods based on traversal or manual experience. The present invention constructs an operating condition frequency offset prediction model, introduces a mapping mechanism between operating condition vectors and frequency offsets, and combines sample set training with variable dimensionality reduction, so that the system can automatically correct extreme frequencies, has strong generalization capabilities, and significantly improves adaptability and stability under different operating conditions. The present invention further calculates the dynamic tolerance bandwidth based on frequency calibration. The present invention further calculates the dynamic tolerance bandwidth based on frequency calibration. ,in Positively correlated with the amplitude of operating condition fluctuations, it has practical physical significance, making the calibrated frequency more robust and engineering feasible. The bandwidth clearly covers the possible range of frequency drift, providing a quantitative basis for subsequent filter design or system matching.

[0113] 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 the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency, characterized in that: include: Real-time collection of high-frequency transformer operating parameters, standardized preprocessing, and integration of preprocessed parameters into a unified format operating condition vector; Defining a frequency response range of the high-frequency transformer, constructing a frequency search space based on the frequency response range, and establishing an objective function for fitness evaluation based on the frequency search space; Using a quantum evolutionary algorithm to search for extreme frequency points in the frequency search space; Based on the operating condition vector and the extreme frequency point, a working condition frequency offset prediction model is constructed, and the working condition frequency offset prediction model outputs the extreme frequency shift offset; A dynamic tolerance bandwidth is calculated according to the predicted offset and the fluctuation range of the operating condition vector.

2. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 1, characterized in that: The operating condition vector is: in, is the input voltage, is the input current, is the coil temperature, is the core temperature, In load state.

3. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 1, characterized in that: Defining the frequency response range of the high-frequency transformer includes the following steps: Calculate the theoretical resonant frequency of a high-frequency transformer based on its equivalent topological parameters The initial frequency range is extended by ±20% with this frequency as the center. ; Based on the initial frequency range, a network analyzer is used to perform a frequency sweep test on the high-frequency transformer to obtain the reflection parameters. , and extract The frequency range is the final frequency response range .

4. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 3, characterized in that: Get the reflection parameters , and extract The specific steps of using the frequency interval as the final frequency response range are: Use a network analyzer to connect the high-frequency transformer to the test port. During the test, one end receives the excitation signal and the other end is connected to the reflection signal measurement channel. The step frequency is set within the initial frequency range to reflect the parameters of the high frequency transformer. Scan measurement, recording the reflection coefficient at each frequency point value; The reflection parameters in the test results Plot or store as frequency- Comparison table, for all frequency points The values ​​are traversed and the filter meets the All frequency point intervals; to satisfy The frequency band of the condition is the boundary, and the continuous frequency interval is extracted as the final frequency response range .

5. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 4, characterized in that: Constructing a frequency search space based on the frequency response range includes the following steps: The frequency step interval Δf is set according to the actual frequency sweep capability of the network analyzer and the response characteristics of the transformer; The final frequency response range obtained by network analyzer screening , the entire frequency band is linearly divided with the set step interval Δf to obtain a set of frequency points arranged in ascending order ,in , , is the number of sampling frequencies; The frequency point set F is used as a basic search unit in the frequency search space to form a one-dimensional continuous search space.

6. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 5, characterized in that: The specific steps of establishing the objective function for fitness evaluation based on the frequency search space are: In the constructed frequency search space, for each frequency sampling point, two key indicator values ​​reflecting the transmission performance of the high-frequency transformer, namely, transmission gain and input return loss, are extracted as a basis for subsequent evaluation; Normalize or standardize the transmission gain value and reflection loss value of each frequency point so that the two parameters are in the same numerical range; Based on the normalized results of transmission gain and reflection loss, a comprehensive transmission characteristic discriminant function is designed as the objective function. The output value of the objective function is used as the fitness evaluation value of the current frequency point; The mathematical expression of the objective function is: in: is the frequency point currently evaluated, obtained discretely in the frequency search space; The signal transmission gain from port 1 input to port 2 output is measured by a network analyzer at the frequency Measured at, it represents the transmission performance of the high-frequency transformer at this frequency; The reflection coefficient of port 1 input is measured by a network analyzer to represent the matching degree of the input end. The smaller the reflection, the better the matching. is the transmission gain; is the reflection coefficient; , It is an adjustable parameter, which can be set by the user or optimized through experience; is a very small positive number.

7. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 1, characterized in that: The quantum evolutionary algorithm is a quantum annealing algorithm or a quantum genetic algorithm, which searches for extreme frequency points and includes the following steps: In the constructed frequency search space, each candidate frequency point is represented by quantum bit encoding to initialize the quantum individual population; Perform quantum measurement operations on each quantum individual in the initialized quantum individual population to collapse the quantum bit sequence of each quantum individual into the corresponding classical bit string, and decode and map the obtained classical bit string to obtain the corresponding frequency sampling points, and then calculate its fitness value in combination with the objective function; According to the size of the fitness value, excellent individuals are selected as the basis for population evolution; Based on the state difference between the current individual and the optimal individual, an adaptive quantum rotating gate is used to update and adjust the amplitude parameters of the quantum bit, guiding the probability distribution of the entire population to evolve towards the global optimal direction; Repeat the quantum measurement, fitness evaluation and quantum gate update operations for multiple rounds of iterative optimization. When the set maximum number of iterations is reached or the fitness convergence of the population meets the termination condition, the search is stopped and the frequency point with the largest fitness value is output as the calibrated extreme frequency point.

8. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 7, characterized in that: The specific steps for using an adaptive quantum rotation gate to update and adjust the amplitude parameters of a quantum bit are as follows: For each quantum individual in the current population, extract the classical bit string obtained from its quantum measurement and compare it bit by bit with the classical bit string of the optimal individual with the highest fitness value in the current iteration. If the bit value on a certain bit is different, it is determined that the bit needs to be rotated and updated; For the bits that need to be rotated and updated, the rotation direction of the quantum rotation gate is set according to the difference between the current bit value and the target bit value, so that the quantum bit amplitude shifts towards the optimal solution direction; and the rotation angle is adaptively determined. . For the quantum bits to be updated, based on the set rotation direction and rotation angle , perform amplitude update operation; Repeat the above rotation update operation to complete the amplitude parameter update of all quantum individuals in the current quantum population and obtain a new generation of quantum individual population.

9. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 1, characterized in that: The specific steps for building the operating frequency offset prediction model are as follows: Under multiple operating conditions, obtain the corresponding operating condition vector And the extreme frequency calibrated under this working condition , with the reference frequency under standard working conditions As a reference, calculate the frequency offset , forming a sample pair ; All the formed sample pairs are combined into a sample set, which is then randomly divided into a training set and a validation set. The training set is used for model construction and parameter fitting, and the validation set is used to evaluate the model prediction accuracy. Standardize the parameters in the operating condition vector and use principal component analysis or correlation analysis to reduce the dimension of variables; A polynomial regression model is constructed to express the nonlinear mapping relationship between the frequency offset and the operating condition vector. The model parameters are solved using the least squares method, and the prediction accuracy is evaluated on the validation set.

10. The method for rapid calibration and tolerance analysis of high-frequency transformer transmission extreme frequency according to claim 1, characterized in that: The calculation formula of the dynamic tolerance bandwidth is: in: 、 The value of is positively correlated with the fluctuation amplitude of the working condition. The downward tolerance indicates the maximum allowable frequency deviation of the transformer under the current working condition. The upward tolerance indicates the maximum allowable frequency deviation value for upward deviation; is the corrected extreme frequency, .

Images