A method and related device for active cancellation of cross interference of charging modules

Abstract

This application discloses an active filtering method and related apparatus for eliminating cross-interference in charging modules, which improves the electromagnetic compatibility and stability of charging systems and reduces the impact of cross-interference. The application includes: monitoring the output spectrum of the charging module and decoupling the spectrum to obtain the interference frequency and intensity; constructing a digital twin model based on the interference frequency and intensity; extracting the operating state characteristics of the target module and combining these characteristics with the expected coupling path analysis to obtain the coupling strength coefficient; generating the expected coupling interference signal generated by the target module according to the digital twin model; generating an anti-vector signal based on the expected coupling interference signal; injecting the anti-vector signal into the pulse width modulation carrier of the target module to generate the actual driving signal; monitoring the residual interference spectrum output by the target module and calculating the deviation between the residual interference spectrum and a preset threshold; and updating the parasitic parameters of the digital twin model and the anti-vector signal based on the deviation.

Description

Technical Field

[0001] This application relates to the fields of power electronics and electromagnetic compatibility technology, and in particular to an active filtering method and related apparatus for eliminating cross-interference in charging modules. Background Technology

[0002] With the rapid development of new energy vehicles, energy storage systems, and smart grids, parallel or cascaded topologies of multiple charging modules have become the mainstream design. However, electromagnetic interference between modules is becoming increasingly prominent, posing a significant challenge to the stable operation of the charging modules and the entire system.

[0003] Crosstalk in charging modules primarily originates from the high-frequency switching of power devices. Rapid voltage and current changes generated by devices such as IGBTs and MOSFETs under pulse width modulation can create coupling paths through parasitic capacitance, parasitic inductance, and the common ground loop between modules. This interference not only manifests as excessive conducted disturbance voltage but can also cause crosstalk between modules, and even lead to faults such as coil no-load operation and overheating damage to power devices in the wireless charging system.

[0004] To suppress the aforementioned interference, existing technologies primarily employ passive filtering schemes. Passive filtering relies on LC filter networks, common-mode chokes, and shielding structures to block interference propagation paths through impedance matching. However, passive filtering schemes have significant limitations: First, multi-stage filtering configurations are required to cover wide-bandgap interference, leading to a substantial increase in size and weight. Second, the hysteresis loss and equivalent series resistance loss of the ferrite core increase sharply at high frequencies; when the switching frequency exceeds 500kHz, the filtering efficiency drops by more than 40%, making it difficult to meet the high-frequency application requirements of next-generation wide-bandgap devices. Third, the parameters of passive components are fixed, making them unable to cope with dynamic interference such as sudden load changes and start-stop transients in the charging module. Summary of the Invention

[0005] This application discloses an active filtering method for eliminating cross-interference in charging modules, including:

[0006] By monitoring the output spectrum of the charging module and decoupling the spectrum, the interference frequency and interference intensity can be obtained.

[0007] Based on the interference frequency and the interference intensity, a digital twin model is constructed. The digital twin model is a multi-module coupled digital model that includes the parasitic effects between modules.

[0008] Extract the working state characteristics of the target module, and combine the working state characteristics with the expected coupling path to obtain the coupling strength coefficient;

[0009] The operating state characteristics, the expected coupling path, and the coupling strength coefficient are input into the digital twin model to generate the expected coupling interference signal generated by the target module;

[0010] An anti-vector signal is generated based on the expected coupling interference signal. The anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase, and is used to cancel the expected coupling interference signal on the coupling path.

[0011] The anti-vector signal is injected into the pulse width modulation carrier of the target module to generate the actual driving signal;

[0012] Monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and a preset threshold.

[0013] Based on the deviation value, update the parasitic parameters of the digital twin model and the anti-vector signal.

[0014] Optionally, generating the anti-vector signal based on the expected coupling interference signal includes:

[0015] The time-domain waveform of the expected coupled interference signal is converted into a discrete complex vector sequence in the frequency domain;

[0016] Based on the discrete complex vector sequence and the preset constraints, a frequency domain complex vector sequence is synthesized;

[0017] The frequency domain complex vector sequence is subjected to spectral shaping and transformation to obtain an anti-vector time-domain discrete signal;

[0018] According to the target module, the sampling rate conversion and amplitude normalization are performed on the anti-vector time-domain discrete signal to generate an anti-vector signal.

[0019] Optionally, the step of performing spectral shaping and transformation on the frequency domain complex vector sequence to obtain an anti-vector time-domain discrete signal includes:

[0020] The frequency domain complex vector sequence is spectrally shaped according to a predefined window function to obtain the shaped frequency domain complex vector sequence.

[0021] A fast Fourier transform is performed on the shaped frequency domain complex vector sequence to obtain an anti-vector time-domain discrete signal.

[0022] Optionally, before performing sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal according to the target module to generate the anti-vector signal, the active elimination filtering method further includes:

[0023] Obtain the carrier frequency and resolution of the pulse width modulator of the target module;

[0024] Optionally, the step of performing sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal according to the target module to generate an anti-vector signal includes:

[0025] Based on the carrier frequency and resolution of the pulse width modulator of the target module, the sampling rate conversion and amplitude normalization are performed on the anti-vector time-domain discrete signal to generate an anti-vector signal.

[0026] Optionally, monitoring the residual interference spectrum output by the target module and calculating the deviation between the residual interference spectrum and a preset threshold includes:

[0027] Acquire the current time-domain signal generated by the target module after the anti-vector signal injection;

[0028] Based on the short-time Fourier transform, dynamic spectrum analysis is performed on the current time-domain signal to obtain the residual interference spectrum;

[0029] The residual interference spectrum is compared with a preset multidimensional spectrum threshold to obtain the deviation value.

[0030] Optionally, the multidimensional spectral threshold includes: a peak limit line and an average energy threshold;

[0031] The step of calculating the deviation value by comparing the residual interference spectrum with a preset multidimensional spectrum threshold includes:

[0032] Identify the frequency points in the residual interference spectrum that exceed the peak limit line, and calculate the maximum peak deviation value of the frequency points that exceed the standard line;

[0033] Calculate the power spectral density integral value of the key frequency band of the residual interference spectrum, and calculate the difference between the power spectral density integral value and the average energy threshold to obtain the average energy deviation.

[0034] Optionally, updating the parasitic parameters of the digital twin model and the anti-vector signal based on the deviation value includes:

[0035] If the deviation value meets the triggering condition, the deviation value is backpropagated to the digital twin model to calculate the sensitivity gradient vector of the parasitic parameters in the digital twin model.

[0036] The deviation value is input into the parameter impact assessment model to calculate the weight coefficient of each parasitic parameter in the digital twin model;

[0037] Based on the weighting coefficients, a set of highly sensitive parameters is selected;

[0038] Based on the magnitude and direction of the sensitivity gradient vector, the parasitic parameters in the set of highly sensitive parameters are corrected;

[0039] Based on the corrected parasitic parameters, the digital twin model is adjusted, and the expected coupling interference signal is regenerated;

[0040] The anti-vector signal is updated based on the expected coupling interference signal.

[0041] Secondly, this application provides an active filtering device for eliminating cross-interference in charging modules, comprising:

[0042] A decoupling unit is used to monitor the output spectrum of the charging module and decouple the spectrum to obtain the interference frequency and interference intensity.

[0043] A construction unit is used to construct a digital twin model based on the interference frequency and the interference intensity, wherein the digital twin model is a multi-module coupled digital model that includes inter-module parasitic effects;

[0044] An extraction unit is used to extract the working state features of the target module and combine the working state features with the expected coupling path analysis to obtain the coupling strength coefficient;

[0045] The first processing unit is used to input the working state characteristics, the expected coupling path and the coupling strength coefficient into the digital twin model to generate the expected coupling interference signal generated by the target module.

[0046] The second processing unit is configured to generate an anti-vector signal based on the expected coupling interference signal, wherein the anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase and is used to cancel the expected coupling interference signal on the coupling path;

[0047] The third processing unit is used to inject the anti-vector signal into the pulse width modulation carrier of the target module to generate the actual driving signal;

[0048] The calculation unit is used to monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and a preset threshold.

[0049] An update unit is used to update the parasitic parameters of the digital twin model and the anti-vector signal based on the deviation value.

[0050] Thirdly, embodiments of this application also provide another active filtering device for eliminating cross-interference in charging modules, comprising:

[0051] Processor, memory, input / output units, and bus;

[0052] The processor is connected to memory, input / output units, and a bus;

[0053] The memory stores a program, which the processor calls to execute, such as the first aspect and any optional active elimination filtering method of the first aspect.

[0054] Fourthly, embodiments of this application provide a computer-readable storage medium storing a program that, when executed on a computer, performs the first aspect and any optional active elimination filtering method of the first aspect.

[0055] As can be seen from the above technical solutions, this application has the following advantages:

[0056] By employing spectrum decoupling technology to obtain interference frequency and intensity, crucial data support is provided for subsequent modeling and analysis. A digital twin model incorporating inter-module parasitic effects is constructed, accurately simulating the electromagnetic characteristics of the actual system. Dynamic calibration ensures the model's accuracy. The operating state characteristics of the target module are extracted, and the coupling strength coefficient is analyzed in conjunction with the expected coupling path, enabling a quantitative assessment of the interference source and extent. Based on this information, an expected coupled interference signal is generated, and a corresponding anti-vector signal is further produced. Through precise phase and amplitude control, effective interference cancellation is achieved on the coupling path. The anti-vector signal is injected into the PWM carrier to generate the actual drive signal, ensuring the normal operation of the system. Finally, by monitoring the residual interference spectrum and calculating the deviation value, the parasitic parameters of the digital twin model and the anti-vector signal are dynamically updated, forming a closed-loop control system. This method overcomes the limitations of traditional passive filtering techniques, adapts to complex and changing operating conditions, significantly improves the electromagnetic compatibility and stability of the charging system, reduces the impact of interference on system performance, and provides strong technical support for the development of high-power-density charging technology. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 A schematic diagram of an embodiment of the active cancellation filtering method for cross-interference in the charging module of this application;

[0059] Figure 2 A schematic diagram of an embodiment of the method for generating anti-vector signals according to this application;

[0060] Figure 3 A schematic diagram of an embodiment of the method for generating vector-reactive discrete-time signals according to this application;

[0061] Figure 4A schematic diagram of another embodiment of the method for generating anti-vector signals according to this application;

[0062] Figure 5 This is a schematic diagram of an embodiment of the method for calculating the deviation value in this application;

[0063] Figure 6 This is a schematic diagram of another embodiment of the method for calculating the deviation value in this application;

[0064] Figure 7 A schematic diagram of an embodiment of the method for updating parasitic parameters and anti-vector signals in this application;

[0065] Figure 8 A schematic diagram of an active filtering device for eliminating cross-interference in the charging module of this application;

[0066] Figure 9 Another schematic diagram of the active filtering device for eliminating cross-interference in the charging module of this application. Detailed Implementation

[0067] To address the aforementioned technical problems, this application provides an active filtering method and related apparatus for eliminating cross-interference in charging modules, which improves the electromagnetic compatibility and stability of the charging system and reduces the impact of cross-interference.

[0068] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0069] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0070] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0071] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0072] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0073] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not an embodiment," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0074] The method described in this application does not depend on a specific type of execution subject, and its implementation can be flexibly adapted to various hardware platforms or software environments. For example, the method can be executed by independently operating electronic devices (including but not limited to smartphones, tablets, personal computers, servers, or dedicated computing devices); it can also be completed collaboratively by multiple nodes in a distributed system, where each node can undertake some or all of the computing tasks; furthermore, the method can also be integrated into embedded systems, IoT terminals, or cloud service platforms. The execution subject, in terms of hardware form, includes but is not limited to general-purpose processors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or neural network processors (NPUs); at the software level, it can be deployed through operating systems, virtual machines, containers, or serverless architectures. Those skilled in the art can select appropriate combinations of execution subjects according to actual needs without departing from the core technical scope of this method.

[0075] To enable those skilled in the art to more accurately understand the core concepts and terms involved in this application, and to facilitate understanding of the method steps and unit functions described in subsequent embodiments, the main terms appearing in this application are explained as follows. The following explanations of terms are intended to define and describe the concepts, but do not limit the implementation methods or technical effects of the present invention. Through these explanations, those skilled in the art can clearly identify the meaning of each concept in this application, their interrelationships, and their role in the overall framework of the method, thereby contributing to a more comprehensive understanding of the content of this application.

[0076] Cross-interference of charging modules: refers to the phenomenon in which, in a system of multiple charging modules connected in parallel or cascaded, unexpected frequency components (such as harmonics and stray noise) appear in the output spectrum of the target module due to electromagnetic coupling, parasitic parameters or power loop interactions between modules, and the interference signal propagates to other modules through spatial radiation or conduction paths.

[0077] Digital twin model: specifically refers to a multi-module coupled digital simulation model built based on the physical characteristics of the charging module.

[0078] Inter-module parasitic effects: Simulates non-ideal coupling paths in actual circuits caused by PCB layout, parasitic parameters of capacitors and inductors, etc.

[0079] Expected coupling path: An interference propagation path predefined according to the module topology, used to calculate the interference contribution of the target module to other modules.

[0080] Coupling strength coefficient: refers to a dimensionless parameter that quantifies the degree of interference of the target module's operating state (such as the conduction loss of the switching transistor and the leakage inductance of the transformer) to other modules. It is obtained by analyzing the correlation between the operating state characteristics and historical interference data, and is used to correct the weight allocation of parasitic parameters in the digital twin model.

[0081] Anti-vector signal: It is a compensation signal generated by the digital twin model with the same center frequency as the expected coupled interference signal and a phase difference of 180°±δ (δ is the allowable phase error range of the system).

[0082] Residual interference spectrum: refers to the set of interference frequency components that are not completely canceled after the target module output signal is compensated by anti-vector signal. Its power spectral density (PSD) distribution is used to evaluate the filtering effect.

[0083] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0084] Please see Figure 1 This application provides an active filtering method for eliminating cross-interference in charging modules, comprising:

[0085] 101. Monitor the output spectrum of the charging module and decouple the spectrum to obtain the interference frequency and interference intensity;

[0086] First, the output voltage / current signal of the charging module is monitored in real time, and the sampling rate is set to be no less than twice the expected highest interference frequency. The sampled data (time domain signal) is converted into frequency domain data. Then, with a dynamic threshold of 6dB above the maximum noise floor, the frequency points exceeding the threshold are extracted as candidate interference frequencies, and the power spectral density at each frequency point is calculated as the interference intensity index.

[0087] 102. Based on the interference frequency and interference intensity, a digital twin model is constructed. The digital twin model is a multi-module coupled digital model that includes the parasitic effects between modules.

[0088] The interaction relationship of N parallel charging modules is described by an equivalent circuit topology, where each module is equivalent to a switching power supply model. Parasitic parameters include PCB trace inductance, switching transistor junction capacitance, and mutual inductance between modules. The initial values ​​of parasitic parameters are extracted by measuring the module port impedance using an impedance analyzer, and a dynamic mapping table between parameters and temperature and load rate is established.

[0089] After the model is established, a test step signal is input to the charging module to run the digital twin model synchronously. If the output voltage waveform error exceeds 5%, the parasitic parameters are iteratively adjusted using the gradient descent method until the error converges to within 1%.

[0090] 103. Extract the working state characteristics of the target module, and combine the working state characteristics with the expected coupling path analysis to obtain the coupling strength coefficient;

[0091] After the model is built, the real-time operating status features of the target module are extracted, including the duty cycle of the switching transistor, the switching frequency and the output power. Combined with the expected coupling path defined by the module layout (such as conductive coupling through the 10cm power bus), the coupling strength coefficient is calculated using a multiple linear regression model.

[0092] Multiple linear regression model:

[0093]

[0094] Where α1, α2, and α3 are regression coefficients, β is the bias term, and D is the duty cycle of the switching transistor. For switching frequency, This refers to the output power.

[0095] The regression coefficients and bias terms are obtained through training with historical interference data; for example, when D=0.5, fsw=200kHz, and Pout=100W, Kcouple=0.8 is calculated, which means that the module contributes 80% to the interference of other modules under this state.

[0096] 104. Input the working state characteristics, expected coupling path and coupling strength coefficient into the digital twin model to generate the expected coupling interference signal generated by the target module;

[0097] The operating state characteristics, expected coupling path, and coupling strength coefficient are input into the digital twin model to simulate the propagation process of each interference frequency component in the parasitic loop. Finally, time-domain waveform data containing all interference frequencies is synthesized as the expected coupled interference signal.

[0098] 105. Generate an anti-vector signal based on the expected coupled interference signal. The anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase, used to cancel the expected coupled interference signal on the coupling path.

[0099] By performing Fourier analysis on the expected coupled interference signal, the frequency components of its fundamental frequency and each harmonic are determined, and the frequency components of the anti-vector signal correspond one-to-one with them. In terms of phase, the phase φ of the expected interference signal is measured by a phase comparator, and the phase of the anti-vector signal is set to φ+180°. In terms of amplitude, considering the loss during signal transmission, the amplitude of the anti-vector signal is set to 1.05-1.1 times the amplitude of the expected interference signal to match the amplitude of the interference signal.

[0100] 106. Inject the anti-vector signal into the pulse width modulation carrier of the target module to generate the actual drive signal;

[0101] Through simulation and field testing, the injection point for the anti-vector signal was determined to be the signal link between the output of the PWM carrier generation circuit and the input of the driver chip. The injected anti-vector signal can directly modulate the carrier waveform in the subsequent stage of PWM signal generation without interfering with the normal operation of the preceding control logic. The anti-vector signal injection point is selected at the peak position of the PWM carrier, and signal superposition is achieved by fine-tuning the duty cycle.

[0102] The PWM carrier signal is a triangular wave with the same frequency as the module's switching frequency, and the anti-vector signal is a sine wave. The two signals are linearly superimposed by an adder. The superimposed composite carrier signal is compared with the reference voltage signal output by the current loop by a comparator to generate a PWM drive signal whose pulse width varies with the anti-vector signal, i.e., the actual drive signal.

[0103] 107. Monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and the preset threshold.

[0104] After the target module is injected with an anti-vector signal and operates stably, the spectrum at its output terminal is monitored again to calculate the residual interference intensity (original interference intensity minus the compensated intensity).

[0105] The deviation value δ is used to represent the degree of difference between the residual interference and the preset threshold. It is calculated using a relative deviation method, and the specific calculation process is described in subsequent embodiments. The calculated deviation values ​​are stored according to frequency points to form a deviation value matrix.

[0106] 108. Update the parasitic parameters and anti-vector signal of the digital twin model based on the deviation value.

[0107] The update trigger condition is set to a deviation value δ of no more than 5%. When the deviation value δ at a certain frequency point is greater than 5%, the update process is triggered; when δ is less than or equal to 5%, it is determined that the current interference cancellation effect meets the requirements and no update is performed.

[0108] The updated parasitic parameters are written into the digital twin model to better match the model's coupling characteristics with the actual system. Simultaneously, the expected coupled interference signal is regenerated based on the updated digital twin model, and new anti-vector signal parameters are determined using the same method as in step 105.

[0109] After the anti-vector signal is updated, the residual interference spectrum is monitored again at 5ms interval and the deviation value is calculated. If the new deviation value is ≤5%, the update process ends; if δ is still >5%, the update process of step 108 is repeated until the deviation value meets the requirements.

[0110] In this embodiment, the frequency and intensity of interference are obtained using spectrum decoupling technology, providing crucial data support for subsequent modeling and analysis. A digital twin model incorporating inter-module parasitic effects is constructed, accurately simulating the electromagnetic characteristics of the actual system. Dynamic calibration ensures the model's accuracy. The operating state characteristics of the target module are extracted, and the coupling strength coefficient is analyzed in conjunction with the expected coupling path, enabling a quantitative assessment of the interference source and extent. Based on this information, an expected coupled interference signal is generated, and a corresponding anti-vector signal is further generated. Through precise phase and amplitude control, effective interference cancellation is achieved on the coupling path. The anti-vector signal is injected into the PWM carrier to generate the actual drive signal, ensuring the normal operation of the system. Finally, by monitoring the residual interference spectrum and calculating the deviation value, the parasitic parameters of the digital twin model and the anti-vector signal are dynamically updated, forming a closed-loop control system. This method overcomes the limitations of traditional passive filtering techniques, adapts to complex and changing operating conditions, significantly improves the electromagnetic compatibility and stability of the charging system, reduces the impact of interference on system performance, and provides strong technical support for the development of high-power-density charging technology.

[0111] Please see Figure 2This application provides an embodiment of a method for generating anti-vector signals, comprising:

[0112] 201. Convert the time-domain waveform of the expected coupled interference signal into a discrete complex vector sequence in the frequency domain;

[0113] 202. Based on the discrete complex vector sequence and the preset constraints, synthesize a frequency domain complex vector sequence;

[0114] 203. Perform spectral shaping and transformation on the frequency domain complex vector sequence to obtain an anti-vector time-domain discrete signal;

[0115] 204. Based on the target module, perform sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal to generate an anti-vector signal.

[0116] The Fast Fourier Transform (FFT) algorithm is used to process the time-domain waveform of the expected coupled interference signal. In practice, an appropriate sampling frequency is selected, and the sampling frequency is at least twice the expected highest interference frequency to ensure that the high-frequency components in the signal can be completely captured.

[0117] Specifically, after sampling the time-domain waveform of the expected coupled interference signal, discrete time-domain sample points are obtained. These time-domain sample points are used as input samples for the Fast Fourier Transform algorithm. After calculation, the output is a discrete complex vector sequence in the frequency domain. Each complex vector represents the amplitude and phase information of a specific frequency component, where the real part represents the amplitude of the cosine component and the imaginary part represents the amplitude of the sine component.

[0118] The preset constraints include amplitude constraints and phase constraints. The amplitude constraint ensures that the generated anti-vector signal will not adversely affect the system due to excessive amplitude. The upper limit of the amplitude of each frequency component of the anti-vector signal is set to 1.2 times the amplitude of the corresponding frequency component of the expected coupled interference signal. The phase constraint requires that the anti-vector signal and the expected coupled interference signal be out of phase at each frequency point, i.e., the phase difference is 180°±5°.

[0119] Based on the discrete complex vector sequence, adjustments are made according to the above constraints. For each frequency component, if its amplitude exceeds the set upper limit, it is scaled down to the upper limit value; at the same time, its phase is adjusted to be within the range opposite to the phase of the expected interference signal, and finally a frequency domain complex vector sequence that meets the constraints is synthesized.

[0120] Convolution operations are used to smooth the spectrum, reduce spectral leakage, and improve the spectral purity of vector signals.

[0121] After spectral shaping, the inverse fast Fourier transform (IFFT) algorithm is used to convert the frequency domain complex vector sequence back to the time domain. The number of points in the IFFT is set to be the same as the number of points in the FFT to ensure signal integrity and accuracy. After IFFT calculation, an anti-vector discrete-time signal is obtained.

[0122] For the PWM carrier frequency of the target module, determine the sampling rate conversion ratio. For example, if the PWM carrier frequency of the target module is 50kHz, and the sampling frequency of the anti-vector discrete-time signal is 200kHz, then a 4x downsampling is required. A polyphase filter bank can be selected to implement the sampling rate conversion. This filter bank can effectively reduce aliasing distortion generated during the sampling rate conversion process and ensure signal quality.

[0123] Amplitude normalization is performed to adjust the amplitude range of the anti-vector discrete-time signal to a range compatible with the target module's driving signal. After sampling rate conversion and amplitude normalization, an anti-vector signal that meets the requirements of the target module is generated.

[0124] In this embodiment, the FFT algorithm is used to convert the time-domain waveform of the expected coupled interference signal into a discrete complex vector sequence in the frequency domain, accurately obtaining the amplitude and phase information of each frequency component, providing a detailed data foundation for subsequent processing. Next, a frequency-domain complex vector sequence is synthesized according to preset amplitude and phase constraints, ensuring that the generated anti-vector signal accurately matches the expected coupled interference signal in amplitude and phase, achieving effective cancellation. Spectrum shaping and the IFFT algorithm are then used to convert the signal back to the time domain, obtaining a discrete time-domain anti-vector signal, improving the spectral purity and quality of the signal. Finally, sampling rate conversion and amplitude normalization are performed according to the requirements of the target module, ensuring that the generated anti-vector signal perfectly adapts to the target module, guaranteeing accurate injection and effective operation in practical applications.

[0125] Please see Figure 3 This application provides an embodiment of a method for generating vector-resilient discrete-time signals, comprising:

[0126] 301. Based on a predefined window function, perform spectral shaping on the frequency domain complex vector sequence to obtain the shaped frequency domain complex vector sequence;

[0127] 302. Perform a Fast Fourier Transform on the shaped frequency domain complex vector sequence to obtain an anti-vector time-domain discrete signal.

[0128] The window function can be either the Blackman window or the Hanning window. Taking the Blackman window as an example, the Blackman window has a wide main lobe and low side lobes, which can effectively reduce spectral leakage and improve the accuracy of spectral analysis.

[0129] A predefined Blackman window is convolved with the frequency domain complex vector sequence. During convolution, the window function weights each frequency component of the frequency domain complex vector sequence, adjusting the amplitude of each frequency component according to the frequency response characteristics of the window function. For example, for high-frequency components, the attenuation effect of the window function reduces their amplitude, thereby suppressing high-frequency noise interference to the signal; for the dominant frequency component, its relative amplitude is kept stable to ensure that the main characteristics of the signal are not affected. After the convolution operation, a shaped frequency domain complex vector sequence is obtained, which has smoother spectral characteristics and significantly reduced sidelobe interference.

[0130] The shaped frequency-domain complex vector sequence is input into a Fast Fourier Transform (FFT) algorithm for iterative calculation. During the calculation, the symmetry and periodicity of the rotation factor are fully utilized to reduce redundant calculations and improve computation speed. After FFT calculation, the frequency-domain signal is converted back to the time domain, resulting in an anti-vector discrete-time signal. This signal contains anti-interference information corresponding to the expected coupled interference signal, and its time-domain waveform characteristics accurately reflect the characteristics of the anti-vector signal, providing an effective signal source for subsequent interference cancellation on the coupling path.

[0131] In this embodiment, a window function is used for spectrum shaping. Leveraging its unique frequency response characteristics, this effectively reduces spectral leakage and suppresses high-frequency noise interference, resulting in a cleaner spectrum for the shaped frequency-domain complex vector sequence. This highlights the main characteristics of the signal and lays a solid foundation for subsequent time-domain conversion. Next, an FFT algorithm is used to ensure the accuracy and efficiency of the frequency-to-time domain conversion, quickly obtaining the anti-vector time-domain discrete signal. This signal accurately corresponds to the expected coupled interference signal, effectively eliminating interference precisely along the coupling path during subsequent interference cancellation, significantly improving the handling effect of cross-interference in the charging module.

[0132] Please see Figure 4 This application provides another embodiment of a method for generating anti-vector signals, comprising:

[0133] 401. Obtain the carrier frequency and resolution of the pulse width modulator of the target module;

[0134] 402. Based on the carrier frequency and resolution of the pulse width modulator of the target module, perform sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal to generate an anti-vector signal.

[0135] The configuration register of the pulse width modulator (PWM) in the target module is read via a hardware interface. Taking a common microcontroller as an example, its PWM module typically has dedicated registers for storing carrier frequency-related parameters. In practice, it is crucial to ensure the accuracy of the read register values. This can be achieved by reading multiple values ​​and averaging them to reduce errors.

[0136] The resolution of a PWM depends on the number of bits in the Automatic Reload Register (ARR). The PWM resolution can be determined by consulting the target module's technical documentation or by directly reading the bit width information of the relevant registers.

[0137] Based on the acquired PWM carrier frequency, determine the sampling rate conversion ratio. The sampling rate conversion ratio is the ratio of the original sampling frequency to the target sampling frequency. When the sampling rate conversion ratio > 1, downsampling is required; when the sampling rate conversion ratio < 1, upsampling is required.

[0138] Downsampling can be achieved using a polyphase filter bank. A polyphase filter bank decomposes the main filter into multiple sub-filters, each processing a different phase of the input signal. For example, if the downsampling ratio is M, the input signal is divided into M phases, each phase is filtered by its corresponding sub-filter, and then every M sample points, one sample is selected as the output.

[0139] Upsampling can be achieved using interpolation methods, such as linear interpolation or cubic spline interpolation. Taking linear interpolation as an example, a new sample point is inserted between two known sample points, and its value is a linear combination of the two known sample points. Through multiple interpolation operations, the sampling rate can be increased.

[0140] Based on the PWM resolution, the amplitude range of the anti-vector time-domain discrete signal after sampling rate conversion is normalized to 0 to the resolution. Between 1 and 1. By normalizing the amplitude, the generated anti-vector signal can be matched with the PWM drive signal of the target module in amplitude, ensuring that the anti-vector signal can be accurately injected into the PWM carrier.

[0141] In this embodiment, during the sampling rate conversion stage, both downsampling and upsampling effectively avoid aliasing distortion and the introduction of high-frequency noise, ensuring signal quality stability. The amplitude normalization step is precisely adjusted according to the PWM resolution to perfectly match the amplitude range of the anti-vector signal with the target module's PWM drive signal.

[0142] Please see Figure 5 This application provides an embodiment of a method for calculating deviation values, comprising:

[0143] 501. Acquire the current time-domain signal generated by the target module after anti-vector signal injection;

[0144] 502. Based on the short-time Fourier transform, perform dynamic spectrum analysis on the current time-domain signal to obtain the residual interference spectrum;

[0145] 503. Calculate the deviation value by comparing the residual interference spectrum with the preset multidimensional spectrum threshold.

[0146] Set appropriate sampling time and number of sampling points according to requirements. The sampling time should be long enough to obtain complete signal characteristics, and the number of sampling points should be reasonably selected based on the frequency components of the signal and analysis requirements to meet the requirements of subsequent short-time Fourier transform. After the target module is subjected to anti-vector signal injection, start acquiring the current time-domain signal.

[0147] Short-time Fourier transform (STFT) is a method for converting a time-domain signal into a time-frequency domain signal. It involves windowing the signal in the time domain and then performing a Fourier transform on the signal within each window to obtain the frequency components of the signal at different time intervals.

[0148] The acquired current time-domain signal is input into a short-time Fourier transform algorithm, and the signal is segmented according to a set window function and window length. A Fourier transform is performed on each segment to obtain its spectral information. The spectral information of all segmented signals is arranged in chronological order to form a dynamic spectrum. The dynamic spectrum can visually display the frequency component changes of the signal over different time periods.

[0149] Extracting residual interference spectrum from the dynamic spectrum diagram. Residual interference spectrum refers to the frequency components of the interference signal that still exist in the target module after anti-vector signal injection. The residual interference spectrum can be separated from the dynamic spectrum diagram by setting appropriate thresholds or filtering methods.

[0150] Based on the design requirements, operating environment, and performance indicators of the target module, determine the multidimensional spectral thresholds. These multidimensional spectral thresholds can include constraints across multiple dimensions, such as peak limits and average energy thresholds.

[0151] The extracted residual interference spectrum is compared point-by-point with a preset multidimensional spectrum threshold. For each frequency point, the difference between the amplitude of the residual interference spectrum and the corresponding threshold is calculated to obtain the deviation value for that frequency point. For frequency range thresholds, it is checked whether there are frequency components in the residual interference spectrum that exceed the frequency range. If so, the corresponding deviation information is recorded.

[0152] By comprehensively processing the deviation values ​​of all frequency points, an overall deviation value index is obtained.

[0153] In this embodiment, the current time-domain signal of the target module after vector signal injection is acquired, and dynamic spectrum analysis is performed using short-time Fourier transform to effectively extract the residual interference spectrum. The deviation value is then calculated by comparing the residual interference spectrum with a preset multidimensional spectrum threshold, which accurately quantifies the degree of residual interference in the target module after vector signal injection. This helps to evaluate the effectiveness of vector signal injection resistance and provides a reliable basis for further optimizing vector signal injection parameters and improving the design of the target module.

[0154] Please see Figure 6 This application provides another embodiment of a method for calculating deviation values, comprising:

[0155] Multidimensional spectral thresholds include: peak limit line and average energy threshold;

[0156] The residual interference spectrum is compared with a preset multidimensional spectrum threshold to obtain the deviation value, including:

[0157] 601. Identify the frequency points in the residual interference spectrum that exceed the peak limit line, and calculate the maximum peak deviation value of the frequency points that exceed the standard.

[0158] 602. Calculate the power spectral density integral value of the key frequency band of the residual interference spectrum, and calculate the difference between the power spectral density integral value and the average energy threshold to obtain the average energy deviation.

[0159] Peak limit data is pre-stored in the system in an appropriate format. The peak limit is an upper limit curve set within a specific frequency range to limit the peak amplitude of residual interference spectrum. It is determined manually based on factors such as the normal operating requirements of the target module, its anti-interference capability, and relevant industry standards.

[0160] Traverse each frequency point of the residual interference spectrum and compare the amplitude value of each frequency point with the amplitude value of the corresponding frequency point of the peak limit line. When it is found that the amplitude value of the residual interference spectrum at a certain frequency point is greater than the amplitude value of the frequency point corresponding to the peak limit line, mark that frequency point as an out-of-limit frequency point.

[0161] For all identified out-of-standard frequency points, calculate the peak deviation value for each out-of-standard frequency point, which is the difference between the amplitude of the residual interference spectrum at that out-of-standard frequency point and the amplitude value at the corresponding frequency point of the peak limit line. Find the maximum value among all the peak deviation values ​​of out-of-standard frequency points as the maximum peak deviation value.

[0162] The critical frequency band is the frequency range that is of primary concern during the normal operation of the target module, or the frequency range that is prone to interference. The critical frequency band is determined based on factors such as the operating characteristics of the target module, the expected effectiveness against vector signal injection, and the actual application scenario.

[0163] Data within the critical frequency band is extracted from the residual interference spectrum, including frequency points and corresponding power spectral density values. The integral value of the power spectral density within the critical frequency band is calculated. The difference between the calculated integral value of the power spectral density in the critical frequency band and a pre-set average energy threshold is calculated to obtain the average energy deviation.

[0164] In this embodiment, by identifying the frequency points exceeding the standard and calculating the maximum peak deviation, it is possible to clearly understand when the residual interference exceeds the allowable range in terms of peak amplitude. This allows for the timely detection of frequency points that may cause serious interference to the target module, providing crucial information for targeted adjustment of anti-vector signal injection parameters or improvement of the target module design. By calculating the deviation between the power spectral density integral value and the average energy threshold in the key frequency band, the overall impact of residual interference in the key frequency band can be assessed from an energy perspective, determining whether it is within a reasonable energy range. Combining these two deviation values ​​allows for a more comprehensive and accurate assessment of the degree of residual interference in the target module after anti-vector signal injection.

[0165] Please see Figure 7 This application provides an embodiment of a method for updating parasitic parameters and anti-vector signals, comprising:

[0166] 701. If the deviation value meets the triggering condition, the deviation value is backpropagated to the digital twin model to calculate the sensitivity gradient vector of the parasitic parameters in the digital twin model.

[0167] 702. Input the deviation value into the parameter impact assessment model and calculate the weight coefficient of each parasitic parameter in the digital twin model;

[0168] 703. Based on the weighting coefficients, select the set of highly sensitive parameters;

[0169] 704. Based on the magnitude and direction of the sensitivity gradient vector, parasitic parameters in a set of highly sensitive parameters are corrected;

[0170] 705. Based on the corrected parasitic parameters, adjust the digital twin model and regenerate the expected coupling interference signal;

[0171] 706. Update the anti-vector signal based on the expected coupled interference signal.

[0172] The pre-set update trigger condition is that the deviation value δ is no greater than 5%. When the deviation value δ at a certain frequency point is greater than 5%, the update process is triggered; when δ is less than or equal to 5%, it is determined that the current interference cancellation effect meets the requirements and no update is performed.

[0173] Establish a backpropagation channel between the digital twin model and the deviation value. During backpropagation, record the gradient information of each layer. Based on the gradient information obtained from backpropagation, calculate the sensitivity gradient for each parasitic parameter in the digital twin model. The sensitivity gradient represents the rate of change of the deviation value for that parasitic parameter.

[0174] The sensitivity gradients of all parasitic parameters are combined into a single vector, known as the sensitivity gradient vector. For example, if there are n parasitic parameters p1, p2, ..., pn, and their corresponding sensitivity gradients are g1, g2, ..., gn, then the sensitivity gradient vector G = [g1, g2, ..., gn]T.

[0175] By analyzing a large amount of historical data, the influence of different parasitic parameters on system performance (related to deviation values) is analyzed, and regression models or neural network models are established as parameter impact assessment models. The parameter impact assessment model is pre-existing within the system.

[0176] The deviation values ​​obtained in step 701 are input into the parameter influence assessment model. Based on the input deviation values, the parameter influence assessment model outputs the weight coefficient of each parasitic parameter in the digital twin model. The weight coefficient reflects the importance of the parasitic parameter's influence on the deviation value; the larger the weight, the more significant the parasitic parameter's influence on the deviation value.

[0177] A screening threshold α is set to determine whether a parasitic parameter is a highly sensitive parameter. This threshold can be set according to actual needs and human experience. All parasitic parameters in the digital twin model are iterated through, and the weight coefficient of each parasitic parameter is compared with the screening threshold α. Parasitic parameters with weight coefficients greater than α are selected to form a set of highly sensitive parameters.

[0178] The magnitude of the sensitivity gradient vector indicates how sensitive the deviation value is to the parasitic parameter, while its direction indicates the trend of the parameter's influence on the deviation value. For example, if the sensitivity gradient is positive, it means that increasing the value of the parasitic parameter will increase the deviation value; therefore, to reduce the deviation value, the value of the parasitic parameter should be decreased. Conversely, if the sensitivity gradient is negative, the value of the parasitic parameter should be increased. Simultaneously, the magnitude of the sensitivity gradient determines the correction magnitude.

[0179] For each parasitic parameter in the set of highly sensitive parameters, the following formula is used for correction:

[0180]

[0181] Where pnew is the corrected parasitic parameter value, pold is the original parasitic parameter value, η is the learning rate (controlling the magnitude of the correction), and g is the sensitivity gradient corresponding to the parasitic parameter.

[0182] Update the corrected parasitic parameters in the digital twin model, replacing the original parasitic parameter values. Check the structure and parameters of the digital twin model to ensure it functions correctly. Use the adjusted digital twin model to regenerate the expected coupling interference signal.

[0183] Please see Figure 8This application provides an embodiment of an active filtering device for eliminating cross-interference in charging modules, comprising:

[0184] The decoupling unit 801 is used to monitor the output spectrum of the charging module and decouple the spectrum to obtain the interference frequency and interference intensity.

[0185] Construction unit 802 is used to construct a digital twin model based on interference frequency and interference intensity. The digital twin model is a multi-module coupled digital model that includes parasitic effects between modules.

[0186] Extraction unit 803 is used to extract the working state features of the target module and combine the working state features with the expected coupling path analysis to obtain the coupling strength coefficient;

[0187] The first processing unit 804 is used to input the working state characteristics, expected coupling path and coupling strength coefficient into the digital twin model to generate the expected coupling interference signal generated by the target module.

[0188] The second processing unit 805 is used to generate an anti-vector signal based on the expected coupling interference signal. The anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase and is used to cancel the expected coupling interference signal on the coupling path.

[0189] The third processing unit 806 is used to inject the anti-vector signal into the pulse width modulation carrier of the target module to generate the actual driving signal;

[0190] The calculation unit 807 is used to monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and a preset threshold.

[0191] The update unit 808 is used to update the parasitic parameters and anti-vector signal of the digital twin model according to the deviation value.

[0192] Optionally, the active elimination filter also includes:

[0193] The acquisition unit 809 is used to acquire the carrier frequency and resolution of the pulse width modulator of the target module.

[0194] For detailed implementation methods, please refer to... Figures 1 to 7 Examples are not detailed here.

[0195] Please see Figure 9 This application provides another active filtering device for eliminating cross-interference in charging modules, including:

[0196] Processor 901, memory 902, input / output unit 903, and bus 904.

[0197] The processor 901 is connected to the memory 902, the input / output unit 903, and the bus 904.

[0198] The memory 902 stores a program, and the processor 901 calls the program to execute it, such as... Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 Active elimination filtering method in [the context].

[0199] This application provides a computer-readable storage medium on which a program is stored, and when the program is executed on a computer, it performs the following... Figure 1 , Figure 2 and Figure 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 Active elimination filtering method in [the context].

[0200] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0201] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods 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 coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0202] 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, depending on actual needs.

[0203] 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.

[0204] 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 medium. 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 storage medium 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 described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. An active filtering method for eliminating cross-interference in charging modules, characterized in that, include: By monitoring the output spectrum of the charging module and decoupling the spectrum, the interference frequency and interference intensity can be obtained. Based on the interference frequency and the interference intensity, a digital twin model is constructed. The digital twin model is a multi-module coupled digital model that includes the parasitic effects between modules. Extract the working state characteristics of the target module, and combine the working state characteristics with the expected coupling path to obtain the coupling strength coefficient; The operating state characteristics, the expected coupling path, and the coupling strength coefficient are input into the digital twin model to generate the expected coupling interference signal generated by the target module; An anti-vector signal is generated based on the expected coupling interference signal. The anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase, and is used to cancel the expected coupling interference signal on the coupling path. The anti-vector signal is injected into the pulse width modulation carrier of the target module to generate the actual driving signal; Monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and a preset threshold. Based on the deviation value, update the parasitic parameters of the digital twin model and the anti-vector signal.

2. The active elimination filtering method according to claim 1, characterized in that, The step of generating an anti-vector signal based on the expected coupled interference signal includes: The time-domain waveform of the expected coupled interference signal is converted into a discrete complex vector sequence in the frequency domain; Based on the discrete complex vector sequence and the preset constraints, a frequency domain complex vector sequence is synthesized; The frequency domain complex vector sequence is subjected to spectral shaping and transformation to obtain an anti-vector time-domain discrete signal; According to the target module, the sampling rate conversion and amplitude normalization are performed on the anti-vector time-domain discrete signal to generate an anti-vector signal.

3. The active elimination filtering method according to claim 2, characterized in that, The step of performing spectral shaping and transformation on the frequency domain complex vector sequence to obtain an anti-vector time-domain discrete signal includes: The frequency domain complex vector sequence is spectrally shaped according to a predefined window function to obtain the shaped frequency domain complex vector sequence. The shaped frequency domain complex vector sequence is subjected to inverse fast Fourier transform to obtain an anti-vector time-domain discrete signal.

4. The active elimination filtering method according to claim 2, characterized in that, Before generating the anti-vector signal by performing sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal according to the target module, the active elimination filtering method further includes: Obtain the carrier frequency and resolution of the pulse width modulator of the target module.

5. The active elimination filtering method according to claim 4, characterized in that, The step of performing sampling rate conversion and amplitude normalization on the anti-vector time-domain discrete signal according to the target module to generate an anti-vector signal includes: Based on the carrier frequency and resolution of the pulse width modulator of the target module, the sampling rate conversion and amplitude normalization are performed on the anti-vector time-domain discrete signal to generate an anti-vector signal.

6. The active elimination filtering method according to claim 1, characterized in that, The monitoring of the residual interference spectrum output by the target module and the calculation of the deviation between the residual interference spectrum and a preset threshold include: Acquire the current time-domain signal generated by the target module after the anti-vector signal injection; Based on the short-time Fourier transform, dynamic spectrum analysis is performed on the current time-domain signal to obtain the residual interference spectrum; The residual interference spectrum is compared with a preset multidimensional spectrum threshold to obtain the deviation value.

7. The active elimination filtering method according to claim 6, characterized in that, The multidimensional spectral thresholds include: peak limit lines and average energy thresholds; The step of calculating the deviation value by comparing the residual interference spectrum with a preset multidimensional spectrum threshold includes: Identify the frequency points in the residual interference spectrum that exceed the peak limit line, and calculate the maximum peak deviation value of the frequency points that exceed the standard line; Calculate the power spectral density integral value of the key frequency band of the residual interference spectrum, and calculate the difference between the power spectral density integral value and the average energy threshold to obtain the average energy deviation.

8. The active elimination filtering method according to claim 7, characterized in that, The step of updating the parasitic parameters of the digital twin model and the anti-vector signal based on the deviation value includes: If the deviation value meets the triggering condition, the deviation value is backpropagated to the digital twin model to calculate the sensitivity gradient vector of the parasitic parameters in the digital twin model. The deviation value is input into the parameter impact assessment model to calculate the weight coefficient of each parasitic parameter in the digital twin model; Based on the weighting coefficients, a set of highly sensitive parameters is selected; Based on the magnitude and direction of the sensitivity gradient vector, the parasitic parameters in the set of highly sensitive parameters are corrected. Based on the corrected parasitic parameters, the digital twin model is adjusted, and the expected coupling interference signal is regenerated; The anti-vector signal is updated based on the expected coupling interference signal.

9. An active filtering device for eliminating cross-interference in charging modules, characterized in that, include: A decoupling unit is used to monitor the output spectrum of the charging module and decouple the spectrum to obtain the interference frequency and interference intensity. A construction unit is used to construct a digital twin model based on the interference frequency and the interference intensity, wherein the digital twin model is a multi-module coupled digital model that includes inter-module parasitic effects; An extraction unit is used to extract the working state features of the target module and combine the working state features with the expected coupling path analysis to obtain the coupling strength coefficient; The first processing unit is used to input the working state characteristics, the expected coupling path and the coupling strength coefficient into the digital twin model to generate the expected coupling interference signal generated by the target module. The second processing unit is configured to generate an anti-vector signal based on the expected coupling interference signal, wherein the anti-vector signal is a signal with the same frequency as the interference signal but opposite in phase and is used to cancel the expected coupling interference signal on the coupling path; The third processing unit is used to inject the anti-vector signal into the pulse width modulation carrier of the target module to generate the actual driving signal; The calculation unit is used to monitor the residual interference spectrum output by the target module and calculate the deviation between the residual interference spectrum and a preset threshold. An update unit is used to update the parasitic parameters of the digital twin model and the anti-vector signal based on the deviation value.

10. An active filtering device for eliminating cross-interference in charging modules, characterized in that, include: The processor, memory, input / output unit, and bus are connected to the memory, the input / output unit, and the bus. The memory stores a program, and the processor calls the program to execute the active elimination filtering method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains a program that, when executed on a computer, performs the active elimination filtering method as described in any one of claims 1 to 8.

Images