Waveband parameter-based load-type parallel resonant network capacitor position estimation method

Abstract

The application provides a load-type parallel resonant network capacitor position estimation method based on wave band parameters, and relates to the wireless communication and radio frequency electronic technology fields. When a target frequency to be tuned is received, the system can quickly determine a target wave band to which the target frequency belongs from multiple working wave bands, and calculate the capacitor position of a variable capacitor used for realizing resonance by inputting the target frequency into an interpolation model corresponding to the target wave band. Through the synergistic effect of wave band modeling, special parameter configuration and accurate model matching, the scheme effectively overcomes the interference influence of load impedance on the traditional estimation method, and finally can significantly improve the capacitor position estimation accuracy in the load-type parallel resonant network.

Description

Technical Field

[0001] This application relates to the fields of wireless communication and radio frequency electronics, specifically to a method for estimating the capacitor location of a load-type parallel resonant network based on band parameters. Background Technology

[0002] Parallel resonant networks, as core components of modern wireless communication systems, are widely used in critical devices such as RF power amplifiers, antenna tuners, and impedance matching networks. During the operation of a parallel resonant network, the position adjustment of the variable capacitor directly determines the network's resonant frequency and impedance characteristics, thus affecting the performance of the entire communication system. Capacitor position estimation technology aims to quickly determine the precise control position of the variable capacitor based on the target operating frequency, avoiding the time delay and frequency deviation problems caused by traditional trial-and-error adjustment methods.

[0003] In related technologies, capacitor location estimation methods primarily employ numerical interpolation to establish the correspondence between frequency and capacitor location. This method estimates location by pre-measuring capacitor location data at multiple frequency points and constructing a lookup table or interpolation function. However, when this approach is applied to load-type parallel resonant networks, especially in wireless communication devices with complex frequency and power combination adjustment requirements, its estimation accuracy faces serious challenges. In practical applications, the introduction of load impedance significantly alters the equivalent circuit parameters of the resonant network, causing a shift in the original frequency-capacitor location relationship. Traditional numerical interpolation methods lack theoretical modeling and mathematical compensation mechanisms for load effects, failing to adaptively adjust the estimation results according to different load conditions, ultimately leading to insufficient accuracy in capacitor location estimation. Summary of the Invention

[0004] This application provides a method for estimating the capacitor position in a load-type parallel resonant network based on band parameters, which can improve the accuracy of capacitor position estimation in the load-type parallel resonant network.

[0005] Firstly, this application provides a method for estimating the capacitor location of a load-type parallel resonant network based on band parameters, the method comprising:

[0006] The frequency range of the parallel resonant network is divided into multiple operating bands, and a unique band parameter is assigned to each operating band.

[0007] Acquire the tuned frequency points and corresponding capacitor location data in each working band, and construct interpolation models for each working band based on the band parameters, tuned frequency points and corresponding capacitor location data.

[0008] Receive the target frequency to be tuned and determine the target band to which the target frequency belongs from multiple operating bands;

[0009] By inputting the target frequency into the interpolation model corresponding to the target band, the capacitance position of the variable capacitor used to achieve resonance is calculated.

[0010] By adopting the above technical solution, the frequency range of the parallel resonant network is divided into multiple operating bands, and a unique band parameter is assigned to each operating band. This effectively solves the problem of the impact of inductance differences and load characteristic variations on estimation accuracy in broadband applications. By acquiring the tuned frequency points and corresponding capacitor position data within each operating band, and constructing interpolation models for each operating band based on the band parameters, tuned frequency points, and corresponding capacitor position data, it is ensured that each interpolation model accurately reflects the true relationship between frequency and capacitor position within that band. When the target frequency to be tuned is received, the system can quickly determine the target band from multiple operating bands. By inputting the target frequency into the interpolation model corresponding to the target band, the capacitor position of the variable capacitor used to achieve resonance is calculated. Through the synergistic effect of band-based modeling, dedicated parameter configuration, and accurate model matching, the solution effectively overcomes the interference of load impedance on traditional estimation methods, ultimately significantly improving the capacitor position estimation accuracy in load-type parallel resonant networks.

[0011] Optionally, the frequency range of the parallel resonant network is divided into multiple operating bands, and a unique band parameter is assigned to each operating band. This includes: dividing the frequency range into multiple operating bands that correspond one-to-one with the multiple discrete inductance values ​​of the inductor elements in the parallel resonant network; assigning a preset identifier as a band parameter to each operating band; and decreasing the inductance of the operating band by a fixed value as the band number of the band parameter increases.

[0012] By adopting the above technical solution, based on the multiple discrete inductance values ​​of the inductor elements in the parallel resonant network, the frequency range is divided into multiple operating bands corresponding one-to-one with these discrete inductance values. This achieves a precise mapping relationship between the frequency range and the inductance parameters, ensuring that the inductance within each operating band remains constant, thus avoiding the error interference caused by inductance changes in capacitor position estimation in traditional methods. By assigning a preset identifier as a band parameter to each operating band, a standardized band identification and management mechanism is established, enabling the system to quickly and accurately locate the operating band to which the target frequency belongs.

[0013] Optionally, based on the band parameters and the capacitance position data corresponding to the tuned frequency points, interpolation models are constructed for each working band, including:

[0014] Based on the band number of the band parameters, construct the formula for the initial resonant frequency corresponding to each working band;

[0015] Substituting the linear relationship between capacitor location and capacitance value and the corrected relationship between load impedance into the initial resonant frequency formula, we obtain the capacitor location estimation formula for each operating band.

[0016] Analyze the tuned frequency points and corresponding capacitor position data under each operating band to determine the type of fitting relationship for the operating band.

[0017] Based on the model construction method and capacitor position estimation formula preset by the fitting relationship type, the tuned frequency points and corresponding capacitor position data are processed to construct interpolation models corresponding to each working band.

[0018] By adopting the above technical solution, and substituting the linear relationship between capacitor position and capacitance value and the corrected relationship between load impedance into the initial resonant frequency formula, the capacitor position estimation formula for each operating band is obtained. This achieves mathematical modeling and theoretical compensation for the load effect, effectively solving the technical problem that traditional numerical interpolation methods cannot handle the influence of load impedance. By analyzing the tuned frequency points and corresponding capacitor position data in each operating band, the fitting relationship type of the operating band is determined, ensuring that the interpolation model can accurately reflect the true mathematical relationship characteristics between frequency and capacitor position in each band.

[0019] Optionally, based on the preset model construction method and capacitor position estimation formula according to the fitting relationship type, the tuned frequency points and corresponding capacitor position data are processed to construct interpolation models corresponding to each working band, including:

[0020] When the fitted relationship type is a linear relationship type or the first nonlinear relationship type, the capacitor position estimation formula is linearized to obtain a linear model.

[0021] Based on the linear model, multiple sets of tuned frequency points and corresponding capacitor location data are constructed into a design matrix and an observation vector.

[0022] Intermediate parameters in the linear model are solved by matrix operations using the least squares method, based on the design matrix and observation vectors.

[0023] Based on the intermediate parameters, the proportional coefficient and bias capacitor value in the capacitor position estimation formula are calculated in reverse, and the interpolation model corresponding to each working band is constructed based on the proportional coefficient and bias capacitor value.

[0024] By adopting the above technical solution, based on a linear model, multiple sets of tuned frequency points and corresponding capacitor position data are constructed into a design matrix and an observation vector, establishing a standardized mathematical expression that allows for effective integration of measured data with the theoretical model. Through matrix operations using the least squares method, intermediate parameters in the linear model are solved based on the design matrix and observation vector. A mature numerical optimization method is employed to ensure the accuracy and numerical stability of the parameter solutions, effectively avoiding the arbitrariness and inefficiency of traditional trial-and-error methods. By reverse-engineering the proportional coefficient and bias capacitor value in the capacitor position estimation formula based on the intermediate parameters, and constructing interpolation models corresponding to each operating band based on the proportional coefficient and bias capacitor value, a precise mapping from measured data to theoretical parameters is achieved, ensuring that the interpolation model conforms to both physical laws and actual measurement results.

[0025] Optionally, based on the preset model construction method and capacitor position estimation formula according to the fitting relationship type, the tuned frequency points and corresponding capacitor position data are processed to construct interpolation models corresponding to each working band, including:

[0026] When the fitting relationship type is the second nonlinear relationship type, the structure of the preset machine learning model is defined in the nonlinear form of the capacitance position estimation formula;

[0027] The nonlinear least squares method is used to fit the machine learning model based on the tuned frequency points and corresponding capacitor position data in each working band, and to learn and solve the optimal values ​​of the proportional coefficient and the bias capacitor.

[0028] By substituting the optimal values ​​of the scaling factor and the bias capacitor into the structure of the machine learning model, the interpolation models corresponding to each working band are obtained.

[0029] By employing the aforementioned technical solution and using the nonlinear least squares method, the machine learning model is fitted based on the tuned frequency points and corresponding capacitor position data within each operating band. The optimal values ​​of the proportional gain and bias capacitor are learned and solved, achieving accurate modeling of complex nonlinear relationships and effectively addressing the influence of nonlinear effects caused by load impedance on capacitor position estimation. The application of the nonlinear least squares method enables the system to automatically find the best parameter combination through iterative optimization while maintaining the physical model structure, significantly improving the model's fitting accuracy to actual data. By substituting the optimal values ​​of the proportional gain and bias capacitor into the structure of the machine learning model, interpolation models corresponding to each operating band are obtained, achieving a deep integration of the theoretical physical model and data-driven optimization. This ensures that the interpolation model possesses both clear physical interpretation and accurately reflects the actual resonance characteristics under complex load conditions.

[0030] Optionally, the method also includes:

[0031] Calculate the capacitance position at the frequency boundary point of two adjacent operating bands, and calculate the boundary difference between the two capacitance positions;

[0032] When the boundary difference is greater than the preset smoothness threshold, the weighted average of the two capacitor positions is calculated, and the weighted average is used as the capacitor position at the frequency intersection of the two adjacent working bands.

[0033] By adopting the above technical solution, when the boundary difference exceeds a preset smoothness threshold, a weighted average of the two capacitor positions is calculated, and this weighted average is used as the capacitor position at the frequency boundary between two adjacent working bands. This establishes an automatic boundary smoothing mechanism, ensuring a smooth transition between interpolation models of different working bands at the boundary. This boundary smoothing method effectively eliminates the abrupt change in frequency switching points caused by band-specific modeling, avoids jumps in capacitor positions during working band switching, and ensures the continuity and stability of capacitor position estimation throughout the entire frequency range.

[0034] Secondly, this application provides a load-type parallel resonant network capacitance location estimation system based on band parameters, the system comprising:

[0035] The band division module is used to divide the frequency range of the parallel resonant network into multiple working bands and assign unique band parameters to each working band.

[0036] The interpolation model construction module is used to obtain the tuned frequency points and corresponding capacitor position data in each working band, and construct the interpolation model corresponding to each working band based on the band parameters, tuned frequency points and corresponding capacitor position data.

[0037] The target band determination module is used to receive the target frequency to be tuned and determine the target band to which the target frequency belongs from multiple working bands;

[0038] The calculation module is used to input the target frequency into the interpolation model corresponding to the target band and calculate the capacitor position of the variable capacitor used to achieve resonance.

[0039] Thirdly, this application provides a computer storage medium that stores multiple instructions adapted for loading by a processor and executing any of the methods described above.

[0040] Fourthly, this application provides an electronic device including a processor, a memory, and a transceiver. The memory is used to store instructions, the transceiver is used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform any of the methods described above.

[0041] In summary, the beneficial effects of the technical solution of this application include:

[0042] By adopting the above technical solution, the frequency range of the parallel resonant network is divided into multiple operating bands, and a unique band parameter is assigned to each operating band. This effectively solves the problem of the impact of inductance differences and load characteristic variations on estimation accuracy in broadband applications. By acquiring the tuned frequency points and corresponding capacitor position data within each operating band, and constructing interpolation models for each operating band based on the band parameters, tuned frequency points, and corresponding capacitor position data, it is ensured that each interpolation model accurately reflects the true relationship between frequency and capacitor position within that band. When the target frequency to be tuned is received, the system can quickly determine the target band from multiple operating bands. By inputting the target frequency into the interpolation model corresponding to the target band, the capacitor position of the variable capacitor used to achieve resonance is calculated. Through the synergistic effect of band-based modeling, dedicated parameter configuration, and accurate model matching, the solution effectively overcomes the interference of load impedance on traditional estimation methods, ultimately significantly improving the capacitor position estimation accuracy in load-type parallel resonant networks. Attached Figure Description

[0043] Figure 1 This is a flowchart illustrating a method for estimating the capacitor location of a load-type parallel resonant network based on band parameters, according to an embodiment of this application.

[0044] Figure 2 This is a schematic diagram of a load-type parallel resonant network capacitor position estimation system based on band parameters according to an embodiment of this application;

[0045] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0046] Explanation of reference numerals in the attached drawings: 300, electronic device; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. Detailed Implementation

[0047] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0048] In the description of the embodiments of this application, words such as "illustrative," "for example," or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "illustrative," "for example," or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Rather, the use of words such as "illustrative," "for example," or "for example" is intended to present the relevant concepts in a specific manner.

[0049] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0050] Please see Figure 1 This document presents a flowchart illustrating a method for estimating the capacitor location of a load-type parallel resonant network based on band parameters, as provided in an embodiment of this application. This method can be implemented using a computer program, a microcontroller, or run on a von Neumann architecture-based system for estimating the capacitor location of a load-type parallel resonant network based on band parameters. The computer program can be integrated into the application or run as a standalone utility application. The specific steps of the method for estimating the capacitor location of a load-type parallel resonant network based on band parameters are described in detail below.

[0051] S101: Divide the frequency range of the parallel resonant network into multiple operating bands and assign a unique band parameter to each operating band.

[0052] In this context, a parallel resonant network refers to a resonant circuit composed of inductors, capacitors, and other components connected in parallel, used to achieve impedance matching and signal selection at a specific frequency. The frequency range refers to the frequency interval within which the resonant network can operate normally, typically determined by the system's technical specifications and component parameters. A working band represents several consecutive frequency sub-intervals obtained by dividing the entire frequency range according to specific rules; the circuit characteristics within each band are relatively stable. Band parameters are unique identifiers assigned to each working band, representing its characteristic attributes, such as band number, center frequency, or corresponding inductance value.

[0053] Specifically, when a parallel resonant network needs to operate over a wide frequency range, due to the nonlinear characteristics of the circuit and the frequency dependence of component parameters, a single mathematical model often cannot maintain sufficient accuracy across the entire frequency range. Therefore, band division is required during system initialization or configuration. This step analyzes the discrete inductance value distribution characteristics of the inductor components in the parallel resonant network, dividing the entire operating frequency range into segments according to the variation law of the inductance value. Each band corresponds to a specific inductance value, making the resonant characteristics of the circuit relatively stable and predictable within that band. Simultaneously, each band is assigned an incremental number as a band parameter; these numbers have a clear correspondence with the inductance value, facilitating subsequent model construction and frequency localization. The granularity of band division needs to balance computational complexity and estimation accuracy; overly fine division increases storage overhead, while overly coarse division affects accuracy.

[0054] In some embodiments, frequency band division and parameter allocation can be achieved in various ways: Optionally, a frequency response analysis method can be used to obtain the impedance characteristic curves across the entire frequency range through frequency sweep testing, identify frequency points with large impedance change rates as band boundaries, and then divide the frequency range into several sub-intervals based on these boundaries. Each sub-interval is then assigned a composite parameter containing center frequency and bandwidth information as a band identifier. It is understood that other methods can also be used for band division, such as segmentation methods based on power transmission efficiency or segmentation strategies based on phase characteristics; these are not limited here.

[0055] Based on the above embodiments, as an optional implementation method, in S101, the frequency range of the parallel resonant network is divided into multiple working bands, and a unique band parameter is assigned to each working band. This can be achieved through the following steps S1011-S1012.

[0056] S1011: Based on the multiple discrete inductance values ​​of the inductor in the parallel resonant network, the frequency range is divided into multiple working bands that correspond one-to-one with the multiple discrete inductance values.

[0057] In a parallel resonant network, an inductor is a passive device used to store magnetic field energy. It is typically implemented using coils or inductors, and its inductance value determines the network's resonant characteristics. Discrete inductance values ​​refer to a finite number of fixed values ​​exhibited by the inductor in a practical circuit. These values ​​are usually determined by the inductor's physical structure, the number of turns in the winding, or the switching state, unlike continuously variable ideal inductors. A one-to-one correspondence means that each operating band is uniquely mapped to a specific discrete inductance value, ensuring the accuracy and uniqueness of band division.

[0058] In practice, the first step is to obtain all available discrete inductance values ​​for the inductor. These values ​​are typically obtained from circuit design documents or actual measurements. Then, based on the fundamental theory of parallel resonant networks, the theoretical resonant frequency range corresponding to each discrete inductance value is calculated. Since different inductance values ​​produce different resonant characteristics, the entire operating frequency range needs to be segmented according to the changes in inductance value. Finally, the inductance values ​​are arranged in descending order, and the frequency boundary points between adjacent inductance values ​​are calculated.

[0059] S1012: Assign a preset identifier as a band parameter to each working band. The inductance of the working band decreases by a fixed value as the band number of the band parameter increases.

[0060] The preset identifier refers to a unique marker defined in advance to distinguish different operating bands, usually in the form of consecutive integer numbers for easy program processing and data management. The band number represents the numerical identifier assigned to each operating band, starting from the smallest number and increasing sequentially, reflecting the relative positional relationship between bands. The fixed decreasing value means that the inductance difference between adjacent bands remains constant; this regular variation pattern simplifies the calculation and prediction of inductance values ​​between bands. The inductance value represents the specific numerical value of the discrete inductance corresponding to each operating band, and this value remains constant throughout the entire band.

[0061] In practice, all operating bands are first sorted in descending order of inductance value to ensure logical consistency of the band sequence. Then, starting with the band number, each band is assigned a consecutive integer identifier, forming an ordered sequence of band parameters. During the assignment process, the decreasing inductance pattern is verified, confirming that the inductance difference between adjacent bands remains constant. This fixed-value decreasing design allows the inductance value of any band to be calculated using a simple linear relationship; that is, the corresponding inductance value can be directly derived from the band number. Finally, the reference inductance value and the fixed decrease amount for the first band are recorded. The inductance value of all subsequent bands can be calculated by subtracting the product of the band number and the decrease amount from the reference value.

[0062] S102: Obtain the tuned frequency points and corresponding capacitor position data in each working band, and construct the interpolation model corresponding to each working band based on the band parameters, tuned frequency points and corresponding capacitor position data.

[0063] The tuned frequency points represent specific frequency values ​​that have successfully achieved resonance during system calibration or testing, and these frequency points have accurate correspondences with capacitor positions. Capacitor position data refers to the specific physical position or control parameters of the variable capacitor when resonance is achieved, typically expressed as the number of steps by a stepper motor, voltage values, or a percentage of relative position. The interpolation model represents a mathematical function or machine learning model built based on a finite number of known data points, used to estimate the capacitor position corresponding to unknown frequency points.

[0064] Specifically, this step is performed during the system calibration phase or initial operation. After band division is completed, an accurate mathematical model needs to be established for each band. First, several tuned frequency points and their corresponding capacitor positions within each working band are obtained through experimental measurements or historical data. This data constitutes the basic sample set for model training. Then, based on the theoretical formula of parallel resonant networks, combined with parameters such as inductance and load impedance for each band, the theoretical relationship between capacitor position and frequency is derived. Next, the distribution characteristics of measured data within each band are analyzed to determine whether a linear relationship or a specific nonlinear mode exists. Based on the data characteristics, an appropriate fitting method is selected. For linear relationships, the least squares method is used for parameter estimation; for complex nonlinear relationships, machine learning algorithms are used for model training. Finally, independent interpolation models for each band are obtained, which can accurately estimate the corresponding capacitor positions based on the input frequency.

[0065] In some embodiments, the construction process of the interpolation model can be implemented in several ways: Optionally, a method combining theoretical modeling and experimental correction can be adopted. First, a theoretical model framework for each band is established based on the capacitance position estimation formula of the parallel resonant network. Then, the proportional coefficient and bias parameter in the model are solved by least squares fitting using the data of the tuned frequency points. Finally, the accuracy of the model is verified and the parameters are fine-tuned as needed to ensure that the estimation error of the model in the band is controlled within an acceptable range. Optionally, a data-driven machine learning method can be adopted. The tuned frequency points are used as input features and the capacitance position is used as the target output to construct a neural network or support vector regression model. The model is trained and the parameters are optimized using the training dataset. The model's generalization performance is evaluated using the cross-validation method, and the model with the smallest validation error is selected as the final interpolation model for the band.

[0066] S103: Receive the target frequency to be tuned and determine the target band to which the target frequency belongs from multiple operating bands;

[0067] The target frequency to be tuned represents the specific frequency value received by the system that requires resonant tuning, typically derived from frequency switching commands from upper-layer applications or user-defined operating frequencies. The target band represents the specific operating band containing the target frequency, meaning the target frequency falls within the frequency range of that band. Furthermore, band assignment determination needs to consider band boundary handling to ensure that each frequency uniquely corresponds to a specific band.

[0068] Specifically, this step is executed in real time when the system receives a new frequency tuning request and is a core component of the online processing phase. Once the target frequency is input, the system needs to quickly and accurately locate the operating band to which that frequency belongs, so that the correct interpolation model can be used for subsequent capacitor location calculations. The band determination process typically employs binary search or direct table lookup. First, the target frequency is compared with the frequency boundaries of each band to find bands that satisfy the condition that the lower boundary is less than or equal to the target frequency and the upper boundary is greater than the target frequency. For boundary frequency points, assignment rules need to be predefined, such as using left-closed, right-open intervals or setting overlapping areas. The accuracy of band determination directly affects the accuracy of subsequent capacitor location estimation; therefore, it is necessary to ensure the completeness and mutual exclusivity of band division to avoid frequency omissions or duplicate assignments.

[0069] In some embodiments, the band determination process for the target frequency can be implemented in several ways: Optionally, a band frequency range lookup table can be established, storing the start and end frequencies of each band in an ordered array. Upon receiving the target frequency, a binary search algorithm is used to quickly locate the corresponding band index. The band's assignment is determined by comparing the target frequency with the band boundary values. This method has a logarithmic time complexity and is suitable for scenarios with a large number of bands. Optionally, a direct calculation method based on band parameters can be used. The theoretical resonant frequency is calculated based on the target frequency and the inductance value corresponding to each band. The band whose theoretical value is closest to the target frequency is found as the target band. At the same time, a frequency deviation threshold is set to verify its validity, ensuring that the target frequency is within the effective working range of the band. It is understood that other methods can also be used to determine the band, such as pattern recognition methods based on frequency features or predictive positioning algorithms based on historical usage frequencies; these are not limited here.

[0070] S104: Input the target frequency into the interpolation model corresponding to the target band, and calculate the capacitor position of the variable capacitor used to achieve resonance.

[0071] In this context, the variable capacitor refers to the capacitor used to adjust the resonant frequency in a parallel resonant network. It is typically implemented using varactor diodes, adjustable capacitors, or capacitor arrays. The capacitor position indicates the specific setting parameters of the variable capacitor, used to control the capacitance value and thus adjust the resonant frequency of the network.

[0072] Specifically, this step, as the final execution stage of the entire capacitor position estimation process, is performed immediately after the target band is determined. First, the interpolation model parameters corresponding to the target band are retrieved from the model library, including key parameters such as the proportional gain and bias capacitance value. Then, the target frequency value is substituted into the interpolation model formula for that band for numerical calculation, progressively calculating various values. The calculation process requires handling mathematical operations such as square root operations and division, while considering numerical accuracy and computational stability. For complex nonlinear models, complex calculations such as polynomial calculations or neural network forward propagation may be involved. The final capacitor position value needs to be checked for reasonableness to ensure that the result is within the physical adjustment range of the variable capacitor, and may require quantization to meet the accuracy requirements of the actual control system.

[0073] Based on the above embodiments, as an optional implementation method, this application provides the following solution to the problem of decreasing inductance and discontinuities between operating bands:

[0074] Calculate the capacitance position at the frequency boundary point of two adjacent working bands respectively, and calculate the boundary difference between the two capacitance positions; when the boundary difference is greater than the preset smoothness threshold, calculate the weighted average of the two capacitance positions, and use the weighted average as the capacitance position of the two adjacent working bands at the frequency boundary point.

[0075] In this context, adjacent operating bands refer to two consecutive bands that are right next to each other in the frequency range division. They are not separated on the frequency axis and share a frequency boundary point. The frequency boundary point represents the common frequency boundary between two adjacent operating bands. This frequency point belongs to both the end frequency of the preceding band and the start frequency of the following band. The boundary difference refers to the absolute difference between the capacitance position values ​​calculated using the interpolation models of the two adjacent bands at the same frequency boundary point. This difference reflects the degree of continuity between the different band models at the boundary.

[0076] In practice, the calculation of boundary differences first involves iterating through all adjacent working band pairs to obtain the frequency boundary point values ​​between each pair. Then, the interpolation models corresponding to the preceding and following bands are called respectively, using the frequency boundary points as input parameters to calculate the capacitance positions. Since the two bands use different inductance values ​​and model parameters, even at the same frequency point, different capacitance position outputs will be produced. Next, the absolute difference between these two capacitance position values ​​is calculated as a measure of the discontinuity at that boundary point. This calculation process needs to be repeated for each frequency boundary point to establish a complete boundary difference dataset.

[0077] To determine the capacitor location, the calculated boundary difference is first compared with a preset smoothness threshold to identify frequency boundaries requiring smoothing. For boundary points exceeding the threshold, a weighted average method is used to calculate new capacitor location values. The weighting coefficients are determined based on multiple factors, including the fitting accuracy of the interpolation model for each band, the richness of the training data, and the frequency of band usage. Typically, bands with higher accuracy or more abundant data are assigned greater weights. When calculating the weighted average, the two original capacitor locations are multiplied by their respective weighting coefficients, and the results are summed to obtain the final smoothed capacitor location. This new capacitor location value will serve as the unified output for two adjacent bands at that frequency boundary, replacing the original two discontinuous values. After smoothing, the boundary processing rules for the relevant bands need to be updated to ensure that a unified boundary value is used in subsequent capacitor location estimations.

[0078] In another embodiment of this application, the construction method of the interpolation model in step S102 can be specifically implemented by the following steps S201-S204.

[0079] S201: Construct the initial resonant frequency formula for each working band according to the band number of the band parameters;

[0080] The initial resonant frequency formula is a mathematical expression describing the relationship between frequency and circuit element parameters, established based on the fundamental theory of parallel resonant networks. This formula serves as the theoretical basis for deriving the capacitor location estimation formula. Constructing a mathematical model based on specific parameters for each frequency band requires incorporating the correspondence between band numbers and inductance values ​​into the theoretical formula.

[0081] In practice, the basic relationship between the resonant frequency and the inductance and capacitance is first determined based on the basic theory of parallel resonant networks.

[0082] The basic relation is: ;

[0083] in, This represents the known inductance value corresponding to the i-th operating band. This represents the capacitance value corresponding to the i-th band.

[0084] Then, using the band number information in the band parameters and the rule that the inductance decreases by a fixed value, a mathematical relationship between the band number and the inductance value is established. The inductance decreasing according to the band number is defined as:

[0085] ;

[0086] in, This represents the reference inductance value, i.e., the inductance of the reference band; i represents the band number, a positive integer sequence starting from 1; This represents the decrease in inductance between adjacent bands and is a fixed constant value.

[0087] Next, this relationship is substituted into the basic resonant frequency expression to form an initial resonant frequency formula with the band number as a parameter. During the construction process, the constancy of the inductance value within each band needs to be considered to ensure the formula's validity within the corresponding band range. By determining the reference inductance value and a fixed decrease, a mapping function from band number to inductance value is established, and this mapping relationship is then embedded into the theoretical expression for the resonant frequency.

[0088] S202: Substitute the linear relationship between capacitor position and capacitance value and the corrected relationship between load impedance into the initial resonant frequency formula to obtain the capacitor position estimation formula for each operating band.

[0089] The linear relationship represents the mathematical relationship between the capacitor position parameters and the actual capacitance value, typically described as a linear function converting the position control parameters to the capacitance value. The load impedance correction formula is a mathematical expression that considers the load's influence on the resonant network, used to correct calculation results under ideal resonance conditions to better reflect the actual circuit's operating state. Substitution represents the mathematical operation of introducing these two relationships as variables into the initial resonant frequency formula.

[0090] This step transforms the theoretical resonant frequency formula into a practical capacitor position estimation formula through mathematical derivation. First, a linear transformation relationship between capacitor position and capacitance value is established. This relationship typically includes two parameters: a proportionality coefficient and an offset constant, reflecting the degree of influence of position change on capacitance value and the zero-position offset, respectively.

[0091] The linear relationship between capacitor location and capacitance value is:

[0092] C_i = k_iP_i + b_i;

[0093] Where C_i represents the actual capacitance of the variable capacitor in the i-th band; the capacitor position P_i represents the physical control position parameter of the variable capacitor; k_i represents the proportional coefficient of the i-th operating band; and b_i represents the bias capacitance value of the i-th operating band.

[0094] Variable capacitors adjust their capacitance by changing the distance between electrodes or the effective area, and this physical change process exhibits good linearity within a certain range. By establishing this linear relationship, the entire estimation system can directly output the corresponding capacitor control position according to the target frequency requirement, realizing a complete closed loop from frequency command to hardware control.

[0095] Then, a correction relationship between load impedance and resonant frequency is established. This correction takes into account the influence of the load on the resonance conditions in the actual circuit, and usually involves the ratio of load resistance to load impedance.

[0096] The corrected formula for load impedance is: ;

[0097] Wherein, the load resistance RL represents the purely resistive component of the load connected to the resonant network, the load impedance ZL represents the total impedance magnitude of the load, and the ratio of RL to twice ZL in the correction factor represents the relative proportion of the load resistance in the total load impedance. The magnitude of this ratio directly affects the strength of the correction. When the load is purely resistive, the load resistance equals the load impedance, and the correction factor reaches its minimum value; when the load is purely reactive, the load resistance is zero, the correction factor equals 1, and it degenerates into the ideal resonance condition.

[0098] In practical applications, resonant networks require an external load to drive them, and the presence of the load alters the circuit's equivalent parameters and resonance conditions. The introduction of load impedance transforms the circuit from a purely LC resonant system into a composite network incorporating resistive losses. Real-world resonant network systems must be connected to a load to function; neglecting the load's influence leads to systematic deviations between the estimated results and actual operating conditions. By introducing a load correction term, the calculated resonant frequency more closely approximates real-world operating conditions, providing a more accurate theoretical basis for subsequent capacitor position estimation. This correction formula considers the impact of different load characteristics on the resonant frequency, enabling the estimation method to adapt to various load types and operating scenarios.

[0099] Next, we substitute these two relationships into the initial resonant frequency formula, and through algebraic transformation and equation solving, we rearrange the formula into a form with the capacitor position as the dependent variable and the frequency as the independent variable.

[0100] The formula for estimating capacitor location is:

[0101] P_i(f)=(1 / (k_i*(2πf)²*L_i))*sqrt(1-(R_L / (2*Z_L))²)-(b_i / k_i);

[0102] Where P_i(f) represents the capacitor position that varies with frequency in the i-th operating band, k_i represents the scaling factor in the i-th operating band, f represents the tuned frequency point, L_i represents the known inductance value in the i-th operating band, R_L and Z_L represent the fixed complex resistor and load impedance, respectively, and b_i represents the bias capacitor value in the i-th operating band.

[0103] The first part of the formula is the frequency-dependent term, which consists of the product of a fractional main term and a load correction factor. This part reflects the main law governing the change of capacitor position with frequency. The second part is the constant correction term, which is expressed as the ratio of the bias capacitor value to the proportional coefficient, and is used to compensate for the reference offset of the capacitor position.

[0104] The final capacitor position estimation formula can directly calculate the corresponding capacitor position value based on the input target frequency. This formula takes into account the band-specific inductance value, the linear transformation characteristics of the capacitor position, and the correction effect of the load impedance.

[0105] S203: Analyze the tuned frequency points and corresponding capacitor position data under each working band to determine the fitting relationship type of the working band;

[0106] Among them, the fitting relationship type refers to the category of mathematical model determined based on the distribution characteristics of actual data.

[0107] This step determines the optimal modeling method for each working band through in-depth analysis of measured data. First, data on tuned frequencies and corresponding capacitor locations within each working band are collected to form a complete data sample set. Then, statistical analysis is performed on the data for each band, calculating distribution characteristic parameters, including mean, variance, correlation coefficient, and other statistics. Next, a scatter plot of the data is drawn to visually observe the distribution patterns and trends between frequencies and capacitor locations. Linear regression analysis is used to verify whether the data conforms to a linear relationship, calculating the correlation coefficient of the linear fit and the results of residual analysis. For data with significant nonlinear characteristics, polynomial fitting tests are further conducted to compare the fitting effects of polynomials of different orders. Simultaneously, curvature analysis and rate of change tests are performed to identify nonlinear components and complex variation patterns in the data. Based on the goodness of fit, residual distribution, and cross-validation results, the most suitable fitting relationship type is determined for each working band.

[0108] S204: Based on the model construction method and capacitor position estimation formula preset by the fitting relationship type, process the tuned frequency points and corresponding capacitor position data to construct interpolation models corresponding to each working band.

[0109] Among them, the model construction method refers to the specific modeling method and parameter estimation strategy predefined for different types of fitting relationships, including different technical paths such as least squares method, nonlinear optimization, and machine learning.

[0110] This step involves building specific models based on the preliminary analysis results, establishing accurate interpolation models for each working band. First, based on the determined fitting relationship type for each band, the appropriate model building method and parameter estimation method are selected. For bands with linear relationships, the least squares method is used for parameter fitting. By constructing a design matrix and observation vectors, matrix operations are used to solve for the unknown coefficients in the linear model. Specifically, the capacitance position estimation formula is linearized, converting the nonlinear terms into a linear form, and then a standard linear regression algorithm is applied for parameter estimation. For bands with significant nonlinear relationships, nonlinear least squares or machine learning methods are used for model training. During nonlinear optimization, the objective function is defined in the original form of the capacitance position estimation formula, and gradient descent or other optimization algorithms are used to iteratively solve for the optimal parameter values.

[0111] In the process of constructing the interpolation model, this application adopts different methods to handle different data fitting situations. For linear relationship types and first nonlinear relationship types with low nonlinearity, steps S301-S304 are adopted on the basis of S204.

[0112] S301: When the fitted relationship type is a linear relationship type or the first nonlinear relationship type, the capacitor position estimation formula is linearized to obtain a linear model;

[0113] The type of fitting relationship refers to the classification criteria of the mathematical relationship between capacitor position and frequency, mainly including two categories: linear relationship type and nonlinear relationship type. A linear relationship type indicates that there is a direct proportional relationship between capacitor position and frequency in a certain transformation form, which can be processed through simple algebraic operations. A nonlinear relationship type refers to a relationship where, although the original relationship between capacitor position and frequency exhibits nonlinear characteristics, it can be transformed into a linear form through appropriate mathematical transformations, such as logarithmic transformation, reciprocal transformation, or square root transformation.

[0114] In practice, when the system detects that the fitted relationship is of a linear or first-order nonlinear type, it initiates the linearization process for the capacitor location estimation formula. The core of linearization is identifying the nonlinear components in the capacitor location estimation formula and transforming them into a linear form using appropriate mathematical transformations. Specifically, this involves replacing the frequency square term in the formula with a new independent variable, thus eliminating the nonlinear characteristics of the original formula. For the load correction factor containing the square root term, its value is pre-calculated and treated as a known constant to avoid introducing additional complexity during linearization. During the transformation, the physical meaning of each parameter in the formula remains unchanged, ensuring that the linearized model accurately reflects the essential relationship between capacitor location and frequency.

[0115] S302: Based on a linear model, multiple sets of tuned frequency points and corresponding capacitor position data are constructed into a design matrix and an observation vector;

[0116] The design matrix is ​​a matrix structure used in linear regression analysis to organize the independent variable data. Each row corresponds to a set of observations, and each column corresponds to an independent variable or model parameter. The observation vector is a column vector containing all the observations of the dependent variable, corresponding to the number of rows in the design matrix.

[0117] Based on the linear model structure obtained after linearization, the collected multiple sets of tuned frequency points and their corresponding capacitance position data are reorganized and rearranged according to the requirements of matrix operations. The construction process of the design matrix first determines the number and form of independent variables in the linear model, and then transforms and arranges the frequency values ​​in each set of observation data according to the requirements of the linear model. Specifically, each tuned frequency point is substituted into the linearized model expression, and the coefficient terms multiplied with the parameters to be determined are extracted. These coefficient terms constitute the corresponding rows of the design matrix. For cases involving multiple bands, it is necessary to classify the frequency points according to their respective bands to ensure that data from different bands are correctly represented in the design matrix. The construction of the observation vector is relatively simple; the capacitance position values ​​in each set of data are directly arranged into column vectors in order corresponding to the row order of the design matrix.

[0118] S303: Solve for intermediate parameters in the linear model using matrix operations of the least squares method, based on the design matrix and observation vectors;

[0119] Least squares is a classic parameter estimation method that finds the optimal values ​​of model parameters by minimizing the sum of squared errors between observed values ​​and model predictions. Intermediate parameters refer to the coefficients to be determined in the linearized model. These parameters may not directly correspond to the physical parameters in the original capacitor location estimation formula and need to be obtained through subsequent inverse calculations to obtain the actual proportional gain and bias capacitor values.

[0120] The intermediate parameters are calculated using the standard least squares matrix operation formula.

[0121] ;

[0122] in, This represents the capacitance position value of the i-th operating band at frequency fj. Indicates intermediate parameters. It is a random error term.

[0123] ;

[0124] The observation vector Y is an n-row, 1-column column vector containing the observed capacitance positions at n different frequency points. Each element Pi(fj) represents the actual capacitance position measured at frequency fj. The design matrix X is an n-row, 2-column matrix. The first column contains the reciprocal of the square of each frequency point, 1 / fj², corresponding to the coefficient of Ai in the model. The second column is all 1s, corresponding to the coefficient of the constant term Ci in the model. This matrix structure corresponds perfectly to the standard form of the linear regression model Y = Xβ + ε.

[0125] The specific calculation process first involves transposing the design matrix, then multiplying the transposed design matrix by the original design matrix to obtain a square matrix. Next, the inverse matrix of this square matrix is ​​calculated; this step requires the square matrix to be non-singular, i.e., its determinant is not zero. Then, the product of the transposed design matrix and the observation vector is calculated, resulting in a vector with the same dimension as the intermediate parameters. The final intermediate parameter vector is obtained by multiplying the aforementioned inverse matrix by the product vector. In actual calculations, to improve numerical stability and computational efficiency, matrix factorization techniques are often used instead of direct matrix inversion, such as QR decomposition, singular value decomposition, or Cholesky decomposition. During the calculation process, the condition number of the matrix needs to be monitored. When the condition number is too large, it indicates the existence of collinearity in the data, requiring the use of regularization techniques or increasing the amount of data to improve the stability of the solution.

[0126] S304: Based on the intermediate parameters, the proportional coefficient and bias capacitor value in the capacitor position estimation formula are calculated in reverse, and the interpolation model corresponding to each working band is constructed based on the proportional coefficient and bias capacitor value.

[0127] The proportional gain is a parameter reflecting the sensitivity of capacitance value to changes in position within the linear relationship of capacitor position; different operating frequency bands have different proportional gain values. The bias capacitance value represents the base capacitance when the capacitor position is zero, used to compensate for the fixed bias of the capacitor and for system reference setting.

[0128] In practice, the specific implementation of the reverse calculation depends on the transformation method used in the initial linearization process, requiring parameter restoration through completely opposite mathematical operations. For linearization using variable substitution, the reverse calculation process needs to convert the substituted variables back into the original frequency expression and rearrange the expressions for the proportional coefficient and bias capacitor value. Special attention must be paid to the reversibility of mathematical transformations and the preservation of numerical accuracy during the calculation process to avoid accumulating calculation errors in multiple mathematical operations. After completing the parameter calculations for each band, a dedicated interpolation model is constructed for each working band, combining the corresponding inductance and load parameters. The interpolation model is constructed using the solved proportional coefficient and bias capacitor value as basic parameters to establish a precise mathematical relationship between frequency and capacitor position within that band. Each interpolation model has a clearly defined applicable frequency range and accuracy index; in practical applications, the appropriate interpolation model can be selected for capacitor position calculation based on the band to which the target frequency belongs.

[0129] In the process of constructing the interpolation model, this application adopts different methods to handle different data fitting situations. For the first nonlinear relationship type with a high degree of nonlinearity, steps S401-S403 are adopted on the basis of S204.

[0130] S401: When the fitting relationship type is the second nonlinear relationship type, the structure of the preset machine learning model is defined in the nonlinear form of the capacitance position estimation formula;

[0131] The second type of nonlinear relationship refers to the relationship between capacitor location and frequency that cannot be transformed into a linear form through simple mathematical transformations. It requires maintaining its original nonlinear mathematical structure. This type of relationship typically contains multiple coupled nonlinear terms, such as high-order frequency terms, trigonometric function terms, exponential function terms, or complex forms like product terms of multiple variables. A pre-defined machine learning model refers to a parameterized model structure pre-designed based on known physical principles and mathematical relationships. The functional form of this model is already determined, but its parameters need to be learned and optimized through data-driven methods. The structural definition of a machine learning model includes the model's input variables, output variables, parameter variables, and the mathematical expressions relating them.

[0132] When the system identifies the fitted relationship as belonging to the second nonlinear relationship type, the complete nonlinear form of the capacitance position estimation formula is directly adopted as the basic structure of the machine learning model. Specifically, the original capacitance position estimation formula Pi(f) = (1 / (ki×(2πf)²×Li)) × √(1-(RL / (2×ZL))²) - (bi / ki) is used as the core function expression of the model, where frequency f is the input variable, capacitance position Pi(f) is the output variable, and the proportional gain ki and bias capacitance value bi are the model parameters to be learned. The model structure also needs to include fixed physical parameters, such as inductance Li, load resistance RL, and load impedance ZL, which remain unchanged during model training. To adapt to the differences in characteristics of different operating bands, independent model instances need to be established for each band, each instance having the same function structure but different parameter values.

[0133] S402: The nonlinear least squares method is used to fit the machine learning model based on the tuned frequency points and corresponding capacitor position data in each working band, and to learn and solve the optimal values ​​of the proportional coefficient and the bias capacitor.

[0134] Nonlinear least squares is a numerical optimization method for solving nonlinear regression problems. It iteratively finds the parameter combination that minimizes the sum of squared errors between the model's predicted and observed values. The core idea of ​​this method is to linearize the nonlinear optimization problem around the current parameter point, then solve the optimal solution of the linearized problem as the parameter value for the next iteration, repeating this process until it converges to a global or local optimum.

[0135] In practice, the process of fitting a machine learning model using the nonlinear least squares method involves several key steps. First, it's necessary to collect and organize data on the tuned frequency points and their corresponding capacitance locations within each working band, ensuring the accuracy and completeness of the data. Then, initial guesses for the parameters are set for each working band. The choice of initial values ​​significantly impacts the algorithm's convergence speed and the final result, and is typically based on physical intuition or preliminary experimental results. Next, the objective function is defined, which is the sum of squared errors between the model's predicted values ​​and the actual observed values. The gradient and Hessian matrix of this function with respect to the parameters to be determined need to be calculable or numerically approximated. The Levenberg-Marquardt algorithm is typically chosen for optimization. This algorithm combines the advantages of the Gauss-Newton method and gradient descent, exhibiting good convergence and numerical stability. During iterative optimization, the algorithm automatically adjusts the parameter values ​​to reduce the objective function value, while monitoring the convergence condition to determine if the optimal solution has been reached. To avoid local optima, a multi-starting-point optimization strategy can be employed, starting optimization from multiple different initial parameter values ​​and then selecting the result that yields the minimum objective function value as the final solution.

[0136] S403: Substitute the optimal values ​​of the proportional coefficient and the bias capacitor into the structure of the machine learning model to obtain the interpolation model corresponding to each working band.

[0137] In practice, substituting the optimized proportional coefficients and bias capacitor values ​​into the machine learning model structure is a relatively straightforward process, but requires careful handling of technical details. First, the optimal parameter values ​​for each working band are substituted into the model's function expression according to the correct correspondence, replacing the original undetermined parameter symbols. The substitution process must maintain consistency in the physical units of the parameters to ensure the correctness of the numerical calculations. After parameter substitution, the original parameterized model is transformed into a specific numerical function, which only includes frequency as the independent variable; all other parameters are already determined to specific values. To facilitate subsequent engineering applications, the interpolation model for each band needs to be encapsulated as an independent function object or computation module. These modules have a unified interface format but contain different internal parameters. The encapsulation of the interpolation model also needs to include necessary input validation functions to check whether the input frequency is within the valid range of the band, providing appropriate error messages or boundary handling for inputs outside the range. To improve computational efficiency, the interpolation model can be pre-calculated and optimized, pre-calculating and storing some computational terms that do not depend on the input frequency, reducing the workload of real-time computation.

[0138] The following are system embodiments of this application, which can be used to execute the method embodiments of this application. For details not disclosed in the system embodiments of this application, please refer to the method embodiments of the application.

[0139] Please see Figure 2 This illustration shows a schematic diagram of a load-type parallel resonant network capacitor location estimation system based on band parameters, provided in an exemplary embodiment of this application. The system can be implemented entirely or partially through software, hardware, or a combination of both. The load-type parallel resonant network capacitor location estimation system based on band parameters includes:

[0140] The band division module is used to divide the frequency range of the parallel resonant network into multiple working bands and assign unique band parameters to each working band.

[0141] The interpolation model construction module is used to obtain the tuned frequency points and corresponding capacitor position data in each working band, and construct the interpolation model corresponding to each working band based on the band parameters, tuned frequency points and corresponding capacitor position data.

[0142] The target band determination module is used to receive the target frequency to be tuned and determine the target band to which the target frequency belongs from multiple working bands;

[0143] The calculation module is used to input the target frequency into the interpolation model corresponding to the target band and calculate the capacitor position of the variable capacitor used to achieve resonance.

[0144] Based on the above embodiments, as an optional embodiment, the band division module is also used to divide the frequency range into multiple working bands corresponding one-to-one with the multiple discrete inductance values ​​of the inductor elements in the parallel resonant network; and to assign a preset identifier as a band parameter to each working band, and the inductance of the working band decreases by a fixed value as the band number of the band parameter increases.

[0145] Based on the above embodiments, as an optional embodiment, the interpolation model construction module is further used to construct the initial resonant frequency formula corresponding to each working band according to the band number of the band parameters; substitute the linear relationship between capacitor position and capacitance value and the correction relationship of load impedance into the initial resonant frequency formula to obtain the capacitor position estimation formula for each working band; analyze the tuned frequency points and corresponding capacitor position data under each working band to determine the fitting relationship type of the working band; process the tuned frequency points and corresponding capacitor position data according to the model construction method preset by the fitting relationship type and the capacitor position estimation formula to construct the interpolation model corresponding to each working band.

[0146] Based on the above embodiments, as an optional embodiment, the interpolation model construction module is further used to linearize the capacitor position estimation formula when the fitting relationship type is a linear relationship type or a first nonlinear relationship type, to obtain a linear model; based on the linear model, multiple sets of tuned frequency points and corresponding capacitor position data are constructed into a design matrix and an observation vector; through matrix operations of the least squares method, the intermediate parameters in the linear model are solved according to the design matrix and the observation vector; based on the intermediate parameters, the scaling factor and bias capacitor value in the capacitor position estimation formula are calculated in reverse, and the interpolation model corresponding to each working band is constructed according to the scaling factor and bias capacitor value.

[0147] Based on the above embodiments, as an optional embodiment, the interpolation model construction module is further used to define the structure of the preset machine learning model in the nonlinear form of the capacitor position estimation formula when the fitting relationship type is the second nonlinear relationship type; using the nonlinear least squares method, the machine learning model is fitted according to the tuned frequency points and corresponding capacitor position data in each working band, and the optimal values ​​of the proportional coefficient and the bias capacitor value are learned and solved; the optimal values ​​of the proportional coefficient and the bias capacitor value are substituted into the structure of the machine learning model to obtain the interpolation model corresponding to each working band.

[0148] Based on the above embodiments, as an optional embodiment, the calculation module is further used to calculate the capacitance position of two adjacent working bands at the frequency boundary point, and calculate the boundary difference between the two capacitance positions; when the boundary difference is greater than the preset smoothness threshold, the weighted average of the two capacitance positions is calculated, and the weighted average is used as the capacitance position of the two adjacent working bands at the frequency boundary point.

[0149] This application also provides a computer storage medium that can store multiple instructions. The instructions are adapted to be loaded and executed by a processor as described in the above embodiment of the load-type parallel resonant network capacitor location estimation method based on band parameters. For the specific execution process, please refer to the detailed description of the embodiment, which will not be repeated here.

[0150] Please see Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 As shown, the electronic device 300 may include: at least one processor 301, at least one network interface 304, user interface 303, memory 305, and at least one communication bus 302.

[0151] The communication bus 302 is used to enable communication between these components.

[0152] The user interface 303 may include a display screen and a camera.

[0153] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0154] The processor 301 may include one or more processing cores. The processor 301 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 305, and by calling data stored in the memory 305. Optionally, the processor 301 may be implemented using at least one hardware form of digital signal processing, field-programmable gate array, or programmable logic array. The processor 301 may integrate one or more of the following: a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 301 and may be implemented as a separate chip.

[0155] The memory 305 may include random access memory (RAM) or read-only memory (ROM). Optionally, the memory 305 may include a non-transitory computer-readable medium. The memory 305 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 305 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), instructions for implementing the various method embodiments described above, etc.; the data storage area may store data involved in the various method embodiments described above, etc. Optionally, the memory 305 may also be at least one storage device located remotely from the aforementioned processor 301. Figure 3 As shown, the memory 305, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for estimating the capacitance position of a load-type parallel resonant network based on band parameters.

[0156] exist Figure 3 In the electronic device 300 shown, the user interface 303 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 301 can be used to call the application program stored in the memory 305, which is a method for estimating the capacitance position of a load-type parallel resonant network based on band parameters. When executed by one or more processors, the electronic device performs one or more methods as described in the above embodiments.

[0157] An electronic device readable storage medium stores instructions that, when executed by one or more processors, cause the electronic device to perform one or more methods as described in the above embodiments.

[0158] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0159] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0160] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the shown or discussed mutual couplings or direct couplings or communication connections may be through some service interfaces; indirect couplings or communication connections between apparatuses or units may be electrical or other forms.

[0161] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0162] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0163] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0164] The above are merely exemplary embodiments of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truths. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure.

Claims

1. A method for estimating the capacitance location of a load-type parallel resonant network based on band parameters, characterized in that, The method includes: The frequency range of the parallel resonant network is divided into multiple operating bands, and a unique band parameter is assigned to each operating band. The tuned frequency points and corresponding capacitor position data within each of the working bands are obtained, and based on the band parameters, the tuned frequency points and the corresponding capacitor position data, interpolation models corresponding to each of the working bands are constructed respectively. The construction of interpolation models for each of the operating bands based on the band parameters and the capacitor position data corresponding to the tuned frequency points includes: According to the band number of the band parameters, an initial resonant frequency formula is constructed for each working band; the linear relationship between capacitor position and capacitance value and the correction relationship of load impedance are substituted into the initial resonant frequency formula to obtain the capacitor position estimation formula for each working band; the tuned frequency points and corresponding capacitor position data under each working band are analyzed to determine the fitting relationship type of the working band; the tuned frequency points and corresponding capacitor position data are processed according to the model construction method preset by the fitting relationship type and the capacitor position estimation formula to construct the interpolation model corresponding to each working band. The formula for estimating the capacitor location is: P_i(f)=(1 / (k_i*(2πf)²*L_i))*sqrt(1-(R_L / (2*Z_L))²)-(b_i / k_i); Wherein, P_i(f) represents the capacitor position that varies with frequency in the i-th operating band, k_i represents the scaling factor of the i-th operating band, f represents the tuned frequency point, L_i represents the known inductance value of the i-th operating band, R_L and Z_L represent the fixed complex resistor and load impedance, respectively, and b_i represents the bias capacitor value of the i-th operating band. Receive the target frequency to be tuned, and determine the target band to which the target frequency belongs from the plurality of operating bands; The target frequency is input into the interpolation model corresponding to the target band to calculate the capacitor position of the variable capacitor used to achieve resonance.

2. The method according to claim 1, characterized in that, The process of dividing the frequency range of the parallel resonant network into multiple operating bands and assigning unique band parameters to each operating band includes: Based on the multiple discrete inductance values ​​of the inductor elements in the parallel resonant network, the frequency range is divided into multiple operating bands that correspond one-to-one with the multiple discrete inductance values. Each of the operating bands is assigned a preset identifier as a band parameter, and the inductance of the operating band decreases by a fixed value as the band number of the band parameter increases.

3. The method according to claim 1, characterized in that, The step of processing the tuned frequency points and the corresponding capacitor position data according to the model construction method preset by the fitting relationship type and the capacitor position estimation formula, and constructing the interpolation model corresponding to each of the working bands, includes: When the fitting relationship type is a linear relationship type or a first nonlinear relationship type, the capacitor position estimation formula is linearized to obtain a linear model. Based on the linear model, multiple sets of tuned frequency points and corresponding capacitor position data are constructed into a design matrix and an observation vector; The intermediate parameters in the linear model are solved by matrix operations using the least squares method, based on the design matrix and the observation vector. Based on the intermediate parameters, the proportional coefficient and bias capacitor value in the capacitor position estimation formula are calculated in reverse, and the interpolation model corresponding to each of the working bands is constructed based on the proportional coefficient and the bias capacitor value.

4. The method according to claim 1, characterized in that, The step of processing the tuned frequency points and the corresponding capacitor position data according to the model construction method preset by the fitting relationship type and the capacitor position estimation formula, and constructing the interpolation model corresponding to each of the working bands, includes: When the fitting relationship type is the second nonlinear relationship type, the structure of the preset machine learning model is defined in the nonlinear form of the capacitance position estimation formula; The nonlinear least squares method is used to fit the machine learning model based on the tuned frequency points and the corresponding capacitor position data in each working band, and to learn and solve for the optimal values ​​of the proportional coefficient and the bias capacitor. Substituting the optimal values ​​of the proportional coefficient and the bias capacitor into the structure of the machine learning model, the interpolation models corresponding to each of the working bands are obtained.

5. The method according to claim 1, characterized in that, The method further includes: Calculate the capacitance position at the frequency boundary point of two adjacent operating bands, and calculate the boundary difference between the two capacitance positions; When the boundary difference is greater than the preset smoothness threshold, the weighted average of the two capacitor positions is calculated, and the weighted average is used as the capacitor position at the frequency intersection point of the two adjacent working bands.

6. A load-type parallel resonant network capacitance location estimation system based on band parameters, characterized in that, The system includes: The band division module is used to divide the frequency range of the parallel resonant network into multiple working bands and assign a unique band parameter to each working band. An interpolation model construction module is used to acquire the tuned frequency points and corresponding capacitor position data within each of the operating bands, and to construct interpolation models corresponding to each of the operating bands based on the band parameters, the tuned frequency points, and the corresponding capacitor position data. The construction of interpolation models corresponding to each of the operating bands based on the band parameters and the capacitor position data corresponding to the tuned frequency points includes: constructing an initial resonant frequency formula for each operating band according to the band number of the band parameters; substituting the linear relationship between capacitor position and capacitance value and the correction relationship of load impedance into the initial resonant frequency formula to obtain the capacitor position estimation formula for each operating band; and analyzing the tuned frequency points and corresponding capacitor position data under each operating band to determine the fitting relationship of the operating band. The system type; according to the model construction method preset by the fitting relationship type and the capacitor position estimation formula, the tuned frequency point and the corresponding capacitor position data are processed to construct the interpolation model corresponding to each working band; the capacitor position estimation formula is: P_i(f)=(1 / (k_i*(2πf)²*L_i))*sqrt(1-(R_L / (2*Z_L))²)-(b_i / k_i); where P_i(f) represents the capacitor position that changes with frequency in the i-th working band, k_i represents the scaling factor of the i-th working band, f represents the tuned frequency point, L_i represents the known inductance value of the i-th working band, R_L and Z_L represent the fixed complex resistance and load impedance, respectively, and b_i represents the bias capacitor value of the i-th working band; The target band determination module is used to receive the target frequency to be tuned and determine the target band to which the target frequency belongs from a plurality of the working bands; The calculation module is used to input the target frequency into the interpolation model corresponding to the target band and calculate the capacitor position of the variable capacitor used to achieve resonance.

7. A computer storage medium, characterized in that, The computer storage medium stores a plurality of instructions, which are adapted to be loaded by a processor and executed as described in any one of claims 1 to 5.

8. An electronic device, characterized in that, The device includes a processor, a memory, and a transceiver. The memory is used to store instructions, the transceiver is used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 5.

Images