Tuning control method of distributed microwave filter

Abstract

The invention relates to the technical field of tuning data of filters, provides a tuning control method of a distributed microwave filter, and aims to solve the problem of difficulty in nonlinear relation and multivariable convergence between a debugging model and an actual model in a filter tuning process. The method comprises the steps of data preprocessing, spectrum analysis, error correction, tuning model construction and optimization and the like, a target frequency component and an interference frequency component in a signal are identified through a fast Fourier transform algorithm, a tuning model is optimized through a genetic algorithm, an optimal tuning parameter is calculated, and if the performance of a filter does not reach a target, the tuning parameter is calculated. According to the method, the accuracy of filter tuning control is improved, and the adaptability and the stability of the filter are enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of tuning data of filters, and in particular to a tuning control method of a distributed microwave filter. Background Art

[0002] Distributed microwave filter is a kind of filter based on transmission line, which has high frequency range and excellent quality factor, and is widely used in microwave filtering in high frequency band. It uses transmission line as medium and shows good frequency selection performance and anti-interference ability. Distributed microwave filter can effectively select the frequency of microwave signal, suppress unwanted signals and improve the accurate transmission of required signals through its unique structural design and transmission line characteristics. This kind of filter plays a key role in communication, radar, satellite navigation and other fields, and is an indispensable component in modern electronic systems.

[0003] During the data processing, the distributed microwave filter first receives the microwave signal from the transmission line. The microwave signal contains components of different frequencies, including both target signals and interference signals. The filter accurately separates and filters the signals of different frequencies. Specifically, the filter allows the target frequency signal to pass smoothly, while attenuating or reflecting other unnecessary frequency components. Finally, the signal after filtering is output, achieving effective purification of the original signal and interference suppression, thereby ensuring the accuracy of data transmission.

[0004] Distributed microwave filters have the following technical pain points in actual use. During the use of distributed microwave filters, the tuning control system faces the nonlinear relationship between the debugging model and the actual model, and the convergence difficulty of the debugging algorithm caused by multiple variables in the debugging process, resulting in reduced accuracy of the tuning control of the distributed microwave filters. In order to solve this technical problem, the present invention provides a tuning control method for a distributed microwave filter. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a tuning control method for a distributed microwave filter to solve the problem that the tuning control system of the distributed microwave filter faces a nonlinear relationship between a debugging model and an actual model during use, and the convergence difficulty of a debugging algorithm caused by multiple variables in the debugging process, which causes the tuning control accuracy of the distributed microwave filter to be reduced.

[0006] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows: The present invention provides a tuning control method for a distributed microwave filter, comprising: Step S101, receiving an original microwave signal and a filter model, preprocessing the original microwave signal to obtain a preprocessed original microwave signal, wherein the preprocessing process of the original microwave signal includes enhancing the original microwave signal strength, filtering the original microwave signal, converting the preprocessed original microwave signal into a digital signal, obtaining microwave data to be detected, and performing data integrity detection on the microwave data to be detected using a preset filtering detection model to obtain a microwave data integrity detection result, wherein the data integrity detection result includes data completeness information and data missing information; Step S102, retrieve the microwave data corresponding to the complete data information, use the microwave data corresponding to the complete data information as the first data to be processed, retrieve the microwave data corresponding to the missing data information, use the microwave data corresponding to the missing data information as the second data to be processed, calculate the average value of the first data to be processed and the second data to be processed to obtain a DC component of the signal, the DC component of the signal includes a first DC component and a second DC component, sum the sample values ​​of the first data to be processed and the second data to be processed respectively to obtain a first sample value and a second sample value, subtract the first DC component from the first sample value to obtain first processed data, and subtract the second DC component from the second sample value to obtain second processed data; Step S103, applying a fast Fourier transform algorithm to perform fast Fourier transform on the first processed data and the second processed data respectively, converting the time domain signal into a frequency domain signal, obtaining the first spectrum data and the second spectrum data, identifying the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data, the target frequency component is the main signal expected to be detected, and the interference frequency component includes a noise signal and a clutter signal, and obtaining a spectrum analysis result, the spectrum analysis result includes the frequency and amplitude of the target frequency component and the interference frequency component, if there is an analysis error in the information corresponding to the second spectrum data in the spectrum analysis result, then using a preset microwave data simulation model to perform data simulation on the data information corresponding to the analysis error, and replacing the data corresponding to the analysis error with the simulated data, so as to obtain repaired data; Step S104, according to the spectrum analysis result, the spectrum analysis historical data is obtained, and the spectrum analysis result and the spectrum analysis historical data are used to build a tuning model of the filter, the tuning model of the filter is used to reflect the nonlinear relationship between the debugging model and the actual model, the filter tuning index information and the tuning effect target are received, the filter tuning index information includes the minimum insertion loss and the maximum out-of-band suppression, the tuning model of the filter is optimized by using a genetic algorithm, and the real-time spectrum analysis result is substituted into the optimized tuning model of the filter, and the tuning parameters required for the filter to achieve the best performance are calculated, and the tuning parameters required for the filter to achieve the best performance include the center frequency adjustment amount and the bandwidth adjustment amount of the filter; Step S105, apply the calculated tuning parameters to the filter, perform actual tuning control operations, monitor the performance changes of the filter, monitor whether the tuning effect meets the tuning effect target, if the performance of the filter does not reach the tuning effect target, then proceed to the next step of optimization, perform gradient descent optimization processing on the tuning parameters, obtain optimized tuning parameters, and output the optimized tuning parameters to the filter.

[0007] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S101 comprises: Acquire historical microwave signal data, the historical microwave signal data includes historical normal signals and historical abnormal signals, preprocess the historical microwave signal data and perform feature extraction to obtain signal strength features, signal frequency features, signal phase features and signal waveform features, and use the extracted features and historical microwave signal data to train a neural network model to obtain a filtering detection model; The preprocessed original microwave signal is input into the filter detection model, and the filter detection model predicts the microwave data to be detected to obtain the prediction result, which includes normal data points and abnormal data points; Normal data points are divided into the data complete part, and abnormal data points are divided into the data missing part.

[0008] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S102 comprises: Retrieving microwave data corresponding to the complete data information as first data to be processed, summing all sample values ​​in the first data to be processed, calculating the number of samples of the first data to be processed, obtaining a summed result of the number of samples of the first data to be processed, and dividing the summed result of the number of samples of the first data to be processed by the number of samples to obtain a first DC component; Retrieving microwave data corresponding to the data missing information as second data to be processed, summing all sample values ​​in the second data to be processed, calculating the number of samples of the second data to be processed, and dividing the summation result by the number of samples to obtain a second DC component; Using all sample values ​​in the first data to be processed, summing all sample values ​​in the first data to be processed to obtain a first sample value; All sample values ​​in the second data to be processed are used to sum all sample values ​​in the second data to be processed to obtain a second sample value.

[0009] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S103 comprises: Acquire first spectrum data obtained after applying a fast Fourier transform algorithm, identify the target frequency component in the first spectrum data, and simultaneously identify the interference frequency component to obtain a result of the first spectrum data; Obtaining second spectrum data obtained after applying the fast Fourier transform algorithm, in the second spectrum data, also identifying the target frequency component and the interference frequency component, obtaining the result of the second spectrum data, comparing the result of the first spectrum data with the result of the second spectrum data, and comparing the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data; If there is an analysis error or abnormal data in the second spectrum data, a preset microwave data simulation model is used to perform data simulation for the analysis error in the information corresponding to the second spectrum data in the spectrum analysis result; The simulated data are used to replace the data corresponding to the analysis error, thereby obtaining the repaired data.

[0010] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S104 comprises: Matching the spectrum analysis results and the filter model in a preset knowledge base, obtaining a model suitable for the nonlinear relationship between the debugging model and the actual model as the selected model, and using the selected model to construct a filter tuning model framework, the filter tuning model framework receives the spectrum analysis results as input, and outputs the filter tuning parameters; Initial parameters of the tuning model framework of the receiving filter, including coefficients, weights and biases of the model; The receiving filter tuning index and the effect target of the filter tuning include the minimum insertion loss and the maximum out-of-band suppression. The tuning model of the filter is optimized using a genetic algorithm. The genetic algorithm searches for the optimal solution in the parameter space by simulating natural selection and genetic mechanisms. The optimized filter tuning parameters are obtained from the output of the genetic algorithm. The optimized filter tuning parameters include the center frequency adjustment amount and the bandwidth adjustment amount of the filter.

[0011] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S104 comprises: Design a fitness function to evaluate the performance of the filter under different tuning parameters; determine the genetic algorithm parameters and set the parameters of the genetic algorithm, such as population size, genetic generation, crossover probability, and mutation probability; A group of initial tuning parameters is randomly generated as a population, wherein the initial tuning parameters include a center frequency adjustment amount and a bandwidth adjustment amount of the filter; Use the fitness function to evaluate the fitness value of each individual in the population, and use the tournament selection method to select excellent individuals for reproduction based on the fitness value; Perform crossover operation on the selected individuals to generate new offspring individuals, perform mutation operation on the offspring individuals, introduce new genes, where genes refer to tuning parameter combinations, and repeat the fitness evaluation, selection, crossover and mutation operations until the set genetic generation is reached. Select the individual with the highest fitness value from the final population output by the genetic algorithm as the optimal solution, and extract the center frequency adjustment amount and bandwidth adjustment amount of the filter from the optimal solution as the optimized tuning parameters; Obtain the latest spectrum analysis results from the real-time spectrum analysis system, including the frequency and amplitude information of the target frequency components and interference frequency components; Substitute the real-time spectrum analysis results into the optimized filter tuning model to calculate the tuning parameters required to achieve the best performance of the filter.

[0012] Furthermore, in the tuning control method of the distributed microwave filter of the present invention, step S105 comprises: Receive and set the learning rate, number of iterations and convergence condition parameters of the gradient descent algorithm, use the tuning parameters as the initial value of the gradient descent algorithm, and the tuning parameters include the center frequency adjustment amount and bandwidth adjustment amount of the filter; Accept the optimization objective function, calculate the partial derivative of the optimization objective function with respect to the tuning parameter, and obtain the gradient vector; Get real-time performance data from the filter monitoring system. Real-time performance data controls insertion loss and out-of-band suppression. Subtract the product of the gradient vector and the learning rate from the current tuning parameter value to get the new parameter value; Check whether the new parameter value meets the convergence condition, which includes whether the modulus of the gradient vector is less than the set threshold or whether the number of iterations reaches the set upper limit; If the convergence condition is not met, continue to update the tuning parameters until the convergence condition is met; After the gradient descent algorithm meets the convergence condition, the final tuning parameter value is obtained, and the final tuning parameter value includes the center frequency adjustment amount and the bandwidth adjustment amount of the filter, and the optimized tuning parameter is applied to the filter.

[0013] Beneficial effects of the present invention: The present invention solves the nonlinear relationship problem between the debugging model and the actual model faced by the distributed microwave filter during use, as well as the convergence difficulty caused by multiple variables during the debugging process, through data processing and algorithm optimization. Through data preprocessing, spectrum analysis, error correction, tuning model construction and optimization, the accuracy and reliability of the tuning parameters are improved, thereby improving the accuracy of the filter tuning control.

[0014] The method of the present invention can dynamically adjust the tuning parameters of the filter according to actual conditions to adapt to different working environments and signal characteristics. The tuning parameters are further optimized by the gradient descent algorithm, so that the filter can maintain stable performance output when facing a complex and changeable signal environment. Through precise tuning control, the present invention helps the filter to achieve smaller insertion loss and greater out-of-band suppression, thereby improving the overall performance of the filter. It is of great significance in improving the transmission efficiency of the communication system, reducing noise interference, and enhancing signal quality.

[0015] The present invention provides an automated tuning control method, which reduces the time and cost of manual debugging. At the same time, by real-time monitoring and dynamic adjustment of the filter performance, the maintenance and replacement costs caused by unstable filter performance are reduced. By continuously optimizing and improving the method, the performance and application range of the filter can be further improved.

[0016] In summary, the present invention significantly improves the accuracy and stability of the tuning control of the distributed microwave filter, reduces the debugging and maintenance costs, and promotes the development of the filter technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0018] Figure 1 A schematic diagram of a tuning control method for a distributed microwave filter provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0020] In order to better understand the purpose of the present invention, the present invention is described in further detail below.

[0021] The present invention provides a tuning control method for a distributed microwave filter, comprising: Step S101, receiving an original microwave signal and a filter model, preprocessing the original microwave signal to obtain a preprocessed original microwave signal, wherein the preprocessing process of the original microwave signal includes enhancing the original microwave signal strength, filtering the original microwave signal, converting the preprocessed original microwave signal into a digital signal, obtaining microwave data to be detected, and performing data integrity detection on the microwave data to be detected using a preset filtering detection model to obtain a microwave data integrity detection result, wherein the data integrity detection result includes data completeness information and data missing information; In step S101, the present invention aims at the tuning control problem of the distributed microwave filter, receives the original microwave signal and the filter model as input, and performs a series of preprocessing operations on the original microwave signal to improve the quality and reliability of the signal and provide an accurate data basis for the subsequent tuning control process. The specific preprocessing process includes: Enhance the original microwave signal strength so that the microwave signal will not lose important information due to low strength during transmission and processing. By enhancing the signal strength, the signal's anti-interference ability and transmission stability can be improved.

[0022] The original microwave signal is filtered to remove noise and interference components in the original microwave signal and retain useful signal features. Through filtering, unnecessary fluctuations in the signal can be reduced and the purity and analyzability of the signal can be improved.

[0023] After the above preprocessing is completed, the preprocessed original microwave signal is converted into a digital signal for subsequent digital processing and analysis. This step is called "obtaining microwave data to be detected".

[0024] Use the preset filtering detection model to perform data integrity detection on the microwave data to be detected. The purpose of this step is to improve the integrity and accuracy of the microwave data and provide a reliable data source for subsequent analysis and processing. The data integrity test results include two parts: Data integrity information refers to microwave data that is considered complete, accurate and reliable during the detection process. These data will be directly used in the subsequent spectrum analysis and tuning control process.

[0025] Missing data refers to missing, abnormal or unreliable microwave data found during the detection process. These data need to be further processed or replaced to improve the accuracy and reliability of the entire tuning control process.

[0026] Through the data preprocessing and data integrity detection in step S101, the present invention provides a high-quality and reliable microwave data foundation for subsequent spectrum analysis, tuning model construction and tuning parameter calculation, thereby effectively solving the nonlinear relationship between the debugging model and the actual model faced by the distributed microwave filter during use, and the convergence difficulty caused by multiple variables in the debugging process to the debugging algorithm, thereby improving the accuracy of tuning control.

[0027] Step S102, retrieve the microwave data corresponding to the complete data information, use the microwave data corresponding to the complete data information as the first data to be processed, retrieve the microwave data corresponding to the missing data information, use the microwave data corresponding to the missing data information as the second data to be processed, calculate the average value of the first data to be processed and the second data to be processed to obtain a DC component of the signal, the DC component of the signal includes a first DC component and a second DC component, sum the sample values ​​of the first data to be processed and the second data to be processed respectively to obtain a first sample value and a second sample value, subtract the first DC component from the first sample value to obtain first processed data, and subtract the second DC component from the second sample value to obtain second processed data; In step S102, the present invention further processes the microwave data obtained in step S101, specifically dividing the data into two parts, namely, complete data information and missing data information, and processes them respectively. The following is a detailed processing process: According to the data integrity detection result in step S101, the microwave data is divided into two categories: microwave data corresponding to complete data information (referred to as first data to be processed) and microwave data corresponding to missing data information (referred to as second data to be processed).

[0028] Calculate the DC component, calculate the average value of the first data to be processed (the complete part of the data), and obtain the DC component of this part of the signal, which is recorded as the first DC component. The DC component is a constant part of the signal and does not change with time.

[0029] Similarly, the second data to be processed (data missing or abnormal part) is also averaged to obtain a second DC component.

[0030] The data processing of summing the sample values ​​is to sum all the sample values ​​of the first data to be processed and the second data to be processed respectively to obtain the first sample value and the second sample value. The first DC component is subtracted from the first sample value to obtain the first processed data. This step is to remove the DC component in the signal so as to more accurately analyze the AC part of the signal. Similarly, the second DC component is subtracted from the second sample value to obtain the second processed data.

[0031] Through the processing of step S102, the present invention further classifies the microwave data, removes the DC component in the signal, and obtains purer AC signal data. This provides a more accurate data basis for subsequent spectrum analysis and tuning control. At the same time, the processing of missing or abnormal data also improves the robustness and accuracy of the entire tuning control process.

[0032] Step S103, applying a fast Fourier transform algorithm to perform fast Fourier transform on the first processed data and the second processed data respectively, converting the time domain signal into a frequency domain signal, obtaining the first spectrum data and the second spectrum data, identifying the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data, the target frequency component is the main signal expected to be detected, and the interference frequency component includes a noise signal and a clutter signal, and obtaining a spectrum analysis result, the spectrum analysis result includes the frequency and amplitude of the target frequency component and the interference frequency component, if there is an analysis error in the information corresponding to the second spectrum data in the spectrum analysis result, then using a preset microwave data simulation model to perform data simulation on the data information corresponding to the analysis error, and replacing the data corresponding to the analysis error with the simulated data, so as to obtain repaired data; In step S103, the present invention performs frequency domain analysis on the data processed in step S102 to identify the target frequency component and the interference frequency component in the signal. The following is a detailed processing process: The fast Fourier transform algorithm is applied to transform the first processed data (from the complete part of the data) to convert the time domain signal into the frequency domain signal to obtain the first spectrum data. Similarly, the second processed data (from the missing or abnormal part of the data, but after the DC component is removed and preliminarily processed) is also fast Fourier transformed to obtain the second spectrum data.

[0033] The target frequency component and the interference frequency component are identified in the first spectrum data and the second spectrum data. The target frequency component is the main signal to be detected, while the interference frequency component includes unnecessary components such as noise signals and clutter signals. Through spectrum analysis, the frequency and amplitude information of the target frequency component and the interference frequency component can be obtained, which is an important basis for subsequent tuning control.

[0034] Compare the target frequency components and interference frequency components in the first spectrum data and the second spectrum data to verify the accuracy of the spectrum analysis results. If analysis errors or abnormal data are found in the second spectrum data (for example, incorrect identification due to missing or abnormal data), use the preset microwave data simulation model to simulate these erroneous data information. Replace the data corresponding to the analysis error with the simulated data to obtain the repaired data. This further improves the accuracy and reliability of the spectrum analysis results.

[0035] Through the frequency domain analysis and data repair in step S103, the present invention successfully converts the time domain signal into the frequency domain signal and accurately identifies the target frequency component and the interference frequency component in the signal. At the same time, through data simulation and repair technology, the spectrum analysis error problem caused by missing or abnormal data is solved, providing accurate and reliable spectrum analysis results for subsequent tuning control.

[0036] Step S104, according to the spectrum analysis result, the spectrum analysis historical data is obtained, and the spectrum analysis result and the spectrum analysis historical data are used to build a tuning model of the filter, the tuning model of the filter is used to reflect the nonlinear relationship between the debugging model and the actual model, the filter tuning index information and the tuning effect target are received, the filter tuning index information includes the minimum insertion loss and the maximum out-of-band suppression, the tuning model of the filter is optimized by using a genetic algorithm, and the real-time spectrum analysis result is substituted into the optimized tuning model of the filter, and the tuning parameters required for the filter to achieve the best performance are calculated, and the tuning parameters required for the filter to achieve the best performance include the center frequency adjustment amount and the bandwidth adjustment amount of the filter; In step S104, the present invention constructs and optimizes the tuning model of the filter based on the spectrum analysis result obtained in step S103 to determine the tuning parameters that enable the filter to achieve optimal performance. The following is a detailed processing process: According to the spectrum analysis results, relevant spectrum analysis history data is obtained. These data provide a historical record of the performance of the filter under different conditions. Using the spectrum analysis results and spectrum analysis history data, a tuning model of the filter is constructed to reflect the nonlinear relationship between the debugging model (i.e., theoretical or simulation model) and the actual model (i.e., the filter in actual work).

[0037] The indicators of receive filter tuning include minimum insertion loss (i.e., minimum energy loss when a signal passes through a filter) and maximum out-of-band suppression (i.e., the filter's ability to suppress unwanted signals). At the same time, the effect target of receive tuning is the performance level that the filter is expected to achieve.

[0038] The tuning model of the filter is optimized by applying a genetic algorithm, which is an optimization algorithm that simulates natural selection and genetic mechanisms and can search for the optimal solution in the parameter space. During the optimization process, the genetic algorithm will continue to iterate, and through operations such as selection, crossover, and mutation, it will gradually improve the tuning parameters of the filter to approach the optimal performance.

[0039] Substitute the real-time spectrum analysis results into the optimized filter tuning model, and calculate the tuning parameters required to achieve the best performance of the filter through model calculation. These parameters usually include the center frequency adjustment and bandwidth adjustment of the filter.

[0040] Through the processing of step S104, the present invention successfully constructs the tuning model of the filter and optimizes it through the genetic algorithm. Finally, the tuning parameters required for the filter to achieve the best performance are calculated, which provides an important basis for the subsequent tuning control operation. This process effectively solves the nonlinear relationship between the debugging model and the actual model, as well as the convergence difficulty caused by multiple variables in the debugging process, thereby improving the accuracy of the tuning control of the distributed microwave filter.

[0041] Step S105, apply the calculated tuning parameters to the filter, perform actual tuning control operations, monitor the performance changes of the filter, monitor whether the tuning effect meets the tuning effect target, if the performance of the filter does not reach the tuning effect target, then proceed to the next step of optimization, perform gradient descent optimization processing on the tuning parameters, obtain optimized tuning parameters, and output the optimized tuning parameters to the filter.

[0042] In step S105, the present invention applies the tuning parameters calculated in the previous steps to the actual filter, performs tuning control operations, and monitors the performance changes of the filter to achieve the expected tuning effect. The following is a detailed processing process: The tuning parameters (including the center frequency adjustment amount and bandwidth adjustment amount of the filter) calculated in step S104 are applied to the actual filter. A tuning control operation is performed to adjust the center frequency and bandwidth of the filter so that its performance meets the design requirements.

[0043] After the tuning control operation is completed, the filter performance changes are continuously monitored. This usually includes monitoring the filter's insertion loss, out-of-band suppression and other key indicators. By comparing with the preset tuning effect target, it is evaluated whether the filter performance meets expectations.

[0044] If the performance of the filter does not reach the tuning effect target, that is, the key indicators do not meet the design requirements, further optimization is required. The gradient descent algorithm is used to optimize the tuning parameters. The gradient descent algorithm is an iterative optimization algorithm that minimizes the objective function (i.e., the optimization target) by continuously adjusting the parameter values. During the optimization process, the gradient vector (i.e., the partial derivative of the objective function with respect to the tuning parameters) is calculated based on the real-time performance data of the filter, and the tuning parameters are updated based on the gradient vector and the learning rate.

[0045] The gradient descent iterations are repeated until the convergence condition is met (for example, the magnitude of the gradient vector is less than the set threshold, or the number of iterations reaches the upper limit).

[0046] When the gradient descent algorithm converges, the optimized tuning parameters are obtained. The optimized tuning parameters are output to the filter again for the final tuning control operation. The performance changes of the filter are monitored again to achieve the expected tuning effect.

[0047] Through the processing of step S105, the present invention can realize accurate tuning control of the distributed microwave filter. Even if the performance is not up to standard during the tuning process, the tuning parameters can be further adjusted through the gradient descent optimization algorithm to make the filter achieve the best performance. This process not only improves the accuracy of the tuning control, but also enhances the adaptability and stability of the filter.

[0048] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S101 comprises: Acquire historical microwave signal data, the historical microwave signal data includes historical normal signals and historical abnormal signals, preprocess the historical microwave signal data and perform feature extraction to obtain signal strength features, signal frequency features, signal phase features and signal waveform features, and use the extracted features and historical microwave signal data to train a neural network model to obtain a filtering detection model; The preprocessed original microwave signal is input into the filter detection model, and the filter detection model predicts the microwave data to be detected to obtain the prediction result, which includes normal data points and abnormal data points; Normal data points are divided into the data complete part, and abnormal data points are divided into the data missing part.

[0049] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S102 includes: Retrieving microwave data corresponding to the complete data information as first data to be processed, summing all sample values ​​in the first data to be processed, calculating the number of samples of the first data to be processed, obtaining a summed result of the number of samples of the first data to be processed, and dividing the summed result of the number of samples of the first data to be processed by the number of samples to obtain a first DC component; Retrieving microwave data corresponding to the data missing information as second data to be processed, summing all sample values ​​in the second data to be processed, calculating the number of samples of the second data to be processed, and dividing the summation result by the number of samples to obtain a second DC component; Using all sample values ​​in the first data to be processed, summing all sample values ​​in the first data to be processed to obtain a first sample value; All sample values ​​in the second data to be processed are used to sum all sample values ​​in the second data to be processed to obtain a second sample value.

[0050] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S103 includes: Acquire first spectrum data obtained after applying a fast Fourier transform algorithm, identify the target frequency component in the first spectrum data, and simultaneously identify the interference frequency component to obtain a result of the first spectrum data; Obtaining second spectrum data obtained after applying the fast Fourier transform algorithm, in the second spectrum data, also identifying the target frequency component and the interference frequency component, obtaining the result of the second spectrum data, comparing the result of the first spectrum data with the result of the second spectrum data, and comparing the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data; If there is an analysis error or abnormal data in the second spectrum data, a preset microwave data simulation model is used to perform data simulation for the analysis error in the information corresponding to the second spectrum data in the spectrum analysis result; The simulated data are used to replace the data corresponding to the analysis error, thereby obtaining the repaired data.

[0051] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S104 includes: Matching the spectrum analysis results and the filter model in a preset knowledge base, obtaining a model suitable for the nonlinear relationship between the debugging model and the actual model as the selected model, and using the selected model to construct a filter tuning model framework, the filter tuning model framework receives the spectrum analysis results as input, and outputs the filter tuning parameters; Initial parameters of the tuning model framework of the receiving filter, including coefficients, weights and biases of the model; The receiving filter tuning index and the effect target of the filter tuning include the minimum insertion loss and the maximum out-of-band suppression. The tuning model of the filter is optimized using a genetic algorithm. The genetic algorithm searches for the optimal solution in the parameter space by simulating natural selection and genetic mechanisms. The optimized filter tuning parameters are obtained from the output of the genetic algorithm. The optimized filter tuning parameters include the center frequency adjustment amount and the bandwidth adjustment amount of the filter.

[0052] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S104 includes: Design a fitness function to evaluate the performance of the filter under different tuning parameters; determine the genetic algorithm parameters and set the parameters of the genetic algorithm, such as population size, genetic generation, crossover probability, and mutation probability; A group of initial tuning parameters is randomly generated as a population, wherein the initial tuning parameters include a center frequency adjustment amount and a bandwidth adjustment amount of the filter; Use the fitness function to evaluate the fitness value of each individual in the population, and use the tournament selection method to select excellent individuals for reproduction based on the fitness value; Perform crossover operation on the selected individuals to generate new offspring individuals, perform mutation operation on the offspring individuals, introduce new genes, where genes refer to tuning parameter combinations, and repeat the fitness evaluation, selection, crossover and mutation operations until the set genetic generation is reached. Select the individual with the highest fitness value from the final population output by the genetic algorithm as the optimal solution, and extract the center frequency adjustment amount and bandwidth adjustment amount of the filter from the optimal solution as the optimized tuning parameters; Obtain the latest spectrum analysis results from the real-time spectrum analysis system, including the frequency and amplitude information of the target frequency components and interference frequency components; Substitute the real-time spectrum analysis results into the optimized filter tuning model to calculate the tuning parameters required to achieve the best performance of the filter.

[0053] Specifically, the tuning control method of the distributed microwave filter of the present invention, step S105 includes: Receive and set the learning rate, number of iterations and convergence condition parameters of the gradient descent algorithm, use the tuning parameters as the initial value of the gradient descent algorithm, and the tuning parameters include the center frequency adjustment amount and bandwidth adjustment amount of the filter; Accept the optimization objective function, calculate the partial derivative of the optimization objective function with respect to the tuning parameter, and obtain the gradient vector; Get real-time performance data from the filter monitoring system. Real-time performance data controls insertion loss and out-of-band suppression. Subtract the product of the gradient vector and the learning rate from the current tuning parameter value to get the new parameter value; Check whether the new parameter value meets the convergence condition, which includes whether the modulus of the gradient vector is less than the set threshold or whether the number of iterations reaches the set upper limit; If the convergence condition is not met, continue to update the tuning parameters until the convergence condition is met; After the gradient descent algorithm meets the convergence condition, the final tuning parameter value is obtained, and the final tuning parameter value includes the center frequency adjustment amount and the bandwidth adjustment amount of the filter, and the optimized tuning parameter is applied to the filter.

[0054] The present invention proposes a tuning control method for a distributed microwave filter, aiming to solve the nonlinear relationship between the debugging model and the actual model in the tuning control system, as well as the convergence difficulty caused by multiple variables in the debugging process, thereby improving the accuracy of the tuning control. The specific solution is as follows: The received raw microwave signal is preprocessed, including signal strength enhancement, filtering, and conversion into a digital signal. The microwave data to be detected is tested for data integrity using a preset filtering detection model to distinguish between complete data information and missing data information.

[0055] The microwave data corresponding to the complete data information and the missing data information are processed separately, and the DC components of each are calculated and subtracted from the sample value to obtain the processed data. The fast Fourier transform algorithm is applied to convert the processed data from the time domain to the frequency domain to identify the target frequency components and the interference frequency components.

[0056] If there are analysis errors in the spectrum analysis results, the preset microwave data simulation model is used to perform data simulation and replace the erroneous data to obtain repaired data.

[0057] According to the spectrum analysis results and spectrum analysis historical data, the tuning model of the filter is constructed, which can reflect the nonlinear relationship between the debugging model and the actual model. The tuning model is optimized using a genetic algorithm, and the optimal solution is searched in the parameter space by simulating natural selection and genetic mechanisms to obtain the optimized filter tuning parameters (including center frequency adjustment and bandwidth adjustment).

[0058] The calculated tuning parameters are applied to the filter to perform actual tuning control operations. The performance changes of the filter are monitored. If the performance does not meet the tuning effect target, the tuning parameters are further optimized using the gradient descent algorithm until the performance requirements are met.

[0059] Through the above steps, the present invention effectively solves the nonlinear relationship problem between the debugging model and the actual model faced in the process of distributed microwave filter tuning control, as well as the convergence difficulty caused by multiple variables. Through data preprocessing, integrity detection, spectrum analysis, error correction, tuning model construction and optimization, and precise application of tuning parameters and performance monitoring, the present invention significantly improves the accuracy of distributed microwave filter tuning control.

Claims

1. A tuning control method for a distributed microwave filter, characterized in that: include: Step S101, receiving an original microwave signal and a filter model, preprocessing the original microwave signal to obtain a preprocessed original microwave signal, wherein the preprocessing process of the original microwave signal includes enhancing the original microwave signal strength, filtering the original microwave signal, converting the preprocessed original microwave signal into a digital signal, obtaining microwave data to be detected, and performing data integrity detection on the microwave data to be detected using a preset filtering detection model to obtain a microwave data integrity detection result, wherein the data integrity detection result includes data completeness information and data missing information; Step S102, retrieve the microwave data corresponding to the complete data information, use the microwave data corresponding to the complete data information as the first data to be processed, retrieve the microwave data corresponding to the missing data information, use the microwave data corresponding to the missing data information as the second data to be processed, calculate the average value of the first data to be processed and the second data to be processed to obtain a DC component of the signal, the DC component of the signal includes a first DC component and a second DC component, sum the sample values ​​of the first data to be processed and the second data to be processed respectively to obtain a first sample value and a second sample value, subtract the first DC component from the first sample value to obtain first processed data, and subtract the second DC component from the second sample value to obtain second processed data; Step S103, applying a fast Fourier transform algorithm to perform fast Fourier transform on the first processed data and the second processed data respectively, converting the time domain signal into a frequency domain signal, obtaining the first spectrum data and the second spectrum data, identifying the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data, the target frequency component is the main signal expected to be detected, and the interference frequency component includes a noise signal and a clutter signal, and obtaining a spectrum analysis result, the spectrum analysis result includes the frequency and amplitude of the target frequency component and the interference frequency component, if there is an analysis error in the information corresponding to the second spectrum data in the spectrum analysis result, then using a preset microwave data simulation model to perform data simulation on the data information corresponding to the analysis error, and replacing the data corresponding to the analysis error with the simulated data, so as to obtain repaired data; Step S104, according to the spectrum analysis result, the spectrum analysis historical data is obtained, and the spectrum analysis result and the spectrum analysis historical data are used to build a tuning model of the filter, the tuning model of the filter is used to reflect the nonlinear relationship between the debugging model and the actual model, the filter tuning index information and the tuning effect target are received, the filter tuning index information includes the minimum insertion loss and the maximum out-of-band suppression, the tuning model of the filter is optimized by using a genetic algorithm, and the real-time spectrum analysis result is substituted into the optimized tuning model of the filter, and the tuning parameters required for the filter to achieve the best performance are calculated, and the tuning parameters required for the filter to achieve the best performance include the center frequency adjustment amount and the bandwidth adjustment amount of the filter; Step S105, apply the calculated tuning parameters to the filter, perform actual tuning control operations, monitor the performance changes of the filter, monitor whether the tuning effect meets the tuning effect target, if the performance of the filter does not reach the tuning effect target, then proceed to the next step of optimization, perform gradient descent optimization processing on the tuning parameters, obtain optimized tuning parameters, and output the optimized tuning parameters to the filter.

2. The tuning control method of a distributed microwave filter according to claim 1, characterized in that: The step S101 includes: Acquire historical microwave signal data, the historical microwave signal data includes historical normal signals and historical abnormal signals, preprocess the historical microwave signal data and perform feature extraction to obtain signal strength features, signal frequency features, signal phase features and signal waveform features, and use the extracted features and historical microwave signal data to train a neural network model to obtain a filtering detection model; The preprocessed original microwave signal is input into the filter detection model, and the filter detection model predicts the microwave data to be detected to obtain the prediction result, which includes normal data points and abnormal data points; Normal data points are divided into the data complete part, and abnormal data points are divided into the data missing part.

3. The tuning control method of a distributed microwave filter according to claim 1, characterized in that: The step S102 includes: Retrieving microwave data corresponding to the complete data information as first data to be processed, summing all sample values ​​in the first data to be processed, calculating the number of samples of the first data to be processed, obtaining a summed result of the number of samples of the first data to be processed, and dividing the summed result of the number of samples of the first data to be processed by the number of samples to obtain a first DC component; Retrieving microwave data corresponding to the data missing information as second data to be processed, summing all sample values ​​in the second data to be processed, calculating the number of samples of the second data to be processed, and dividing the summation result by the number of samples to obtain a second DC component; Using all sample values ​​in the first data to be processed, summing all sample values ​​in the first data to be processed to obtain a first sample value; All sample values ​​in the second data to be processed are used to sum all sample values ​​in the second data to be processed to obtain a second sample value.

4. The tuning control method of a distributed microwave filter according to claim 1, characterized in that: The step S103 includes: Acquire first spectrum data obtained after applying a fast Fourier transform algorithm, identify the target frequency component in the first spectrum data, and simultaneously identify the interference frequency component to obtain a result of the first spectrum data; Obtaining second spectrum data obtained after applying the fast Fourier transform algorithm, in the second spectrum data, also identifying the target frequency component and the interference frequency component, obtaining the result of the second spectrum data, comparing the result of the first spectrum data with the result of the second spectrum data, and comparing the target frequency component and the interference frequency component in the first spectrum data and the second spectrum data; If there is an analysis error or abnormal data in the second spectrum data, a preset microwave data simulation model is used to perform data simulation for the analysis error in the information corresponding to the second spectrum data in the spectrum analysis result; The simulated data are used to replace the data corresponding to the analysis error, thereby obtaining the repaired data.

5. The tuning control method of a distributed microwave filter according to claim 1, characterized in that: The step S104 includes: Matching the spectrum analysis results and the filter model in a preset knowledge base, obtaining a model suitable for the nonlinear relationship between the debugging model and the actual model as the selected model, and using the selected model to construct a filter tuning model framework, the filter tuning model framework receives the spectrum analysis results as input, and outputs the filter tuning parameters; Initial parameters of the tuning model framework of the receiving filter, including coefficients, weights and biases of the model; The receiving filter tuning index and the effect target of the filter tuning include the minimum insertion loss and the maximum out-of-band suppression. The tuning model of the filter is optimized using a genetic algorithm. The genetic algorithm searches for the optimal solution in the parameter space by simulating natural selection and genetic mechanisms. The optimized filter tuning parameters are obtained from the output of the genetic algorithm. The optimized filter tuning parameters include the center frequency adjustment amount and the bandwidth adjustment amount of the filter.

6. The tuning control method of the distributed microwave filter according to claim 5, characterized in that: The step S104 includes: Design a fitness function to evaluate the performance of the filter under different tuning parameters; determine the genetic algorithm parameters and set the parameters of the genetic algorithm, such as population size, genetic generation, crossover probability, and mutation probability; A group of initial tuning parameters is randomly generated as a population, wherein the initial tuning parameters include a center frequency adjustment amount and a bandwidth adjustment amount of the filter; Use the fitness function to evaluate the fitness value of each individual in the population, and use the tournament selection method to select excellent individuals for reproduction based on the fitness value; Perform crossover operation on the selected individuals to generate new offspring individuals, perform mutation operation on the offspring individuals, introduce new genes, where genes refer to tuning parameter combinations, and repeat the fitness evaluation, selection, crossover and mutation operations until the set genetic generation is reached. Select the individual with the highest fitness value from the final population output by the genetic algorithm as the optimal solution, and extract the center frequency adjustment amount and bandwidth adjustment amount of the filter from the optimal solution as the optimized tuning parameters; Obtain the latest spectrum analysis results from the real-time spectrum analysis system, including the frequency and amplitude information of the target frequency components and interference frequency components; Substitute the real-time spectrum analysis results into the optimized filter tuning model to calculate the tuning parameters required to achieve the best performance of the filter.

7. The tuning control method of a distributed microwave filter according to claim 1, characterized in that: The step S105 includes: Receive and set the learning rate, number of iterations and convergence condition parameters of the gradient descent algorithm, use the tuning parameters as the initial value of the gradient descent algorithm, and the tuning parameters include the center frequency adjustment amount and bandwidth adjustment amount of the filter; Accept the optimization objective function, calculate the partial derivative of the optimization objective function with respect to the tuning parameter, and obtain the gradient vector; Get real-time performance data from the filter monitoring system. Real-time performance data controls insertion loss and out-of-band suppression. Subtract the product of the gradient vector and the learning rate from the current tuning parameter value to get the new parameter value; Check whether the new parameter value meets the convergence condition, which includes whether the modulus of the gradient vector is less than the set threshold or whether the number of iterations reaches the set upper limit; If the convergence condition is not met, continue to update the tuning parameters until the convergence condition is met; After the gradient descent algorithm meets the convergence condition, the final tuning parameter value is obtained, and the final tuning parameter value includes the center frequency adjustment amount and the bandwidth adjustment amount of the filter, and the optimized tuning parameter is applied to the filter.

Images