Voltage control method, device, equipment and storage medium of pulse power supply

Abstract

The invention relates to the field of voltage control technology, and discloses a voltage control method, device, equipment and storage medium of a pulse power supply. The method comprises: obtaining the input voltage of the pulse power supply and performing multi-phase shunting processing to obtain N voltage sources; constructing an initial pulse frequency modulation model and performing dynamic pulse control to generate N groups of staggered pulse sequences; performing processing to obtain pulse output with low harmonic distortion; performing high-precision voltage sampling to obtain real-time voltage feedback data; co-optimizing the voltage outer loop and the current inner loop to generate a control compensation signal; dynamically adjusting the parameters of the initial pulse frequency modulation model to obtain a target pulse frequency modulation model, and adjusting the output voltage of the pulse power supply through the target pulse frequency modulation model. The invention can realize effective utilization of the input voltage, precise regulation of the output voltage, and effective suppression of harmonic distortion, and at the same time has good dynamic response capability and anti-interference capability.

Description

Technical Field

[0001] The present invention relates to the field of voltage control technology, and in particular to a voltage control method, device, equipment and storage medium for a pulse power supply. Background Art

[0002] Traditional pulse power supply control methods often fail to meet the requirements of high precision, low harmonics, and fast response, especially under complex load and unstable input conditions, where control performance is easily affected. Currently, a single control strategy is difficult to fully solve the various challenges faced by pulse power supplies, such as output voltage fluctuations, harmonic distortion, and slow dynamic response.

[0003] In addition, with the diversification of application scenarios, pulse power supplies need to adapt to more complex working environments, including wide-range input voltage changes, load mutations, grid disturbances, etc. Traditional fixed parameter control methods are difficult to cope with these complex working conditions, and are prone to problems such as reduced system stability and reduced output quality. Summary of the invention

[0004] The present invention provides a voltage control method, device, equipment and storage medium for a pulse power supply. The present invention can achieve effective utilization of input voltage, precise regulation of output voltage, and effective suppression of harmonic distortion, while having good dynamic response capability and anti-interference capability.

[0005] In a first aspect, the present invention provides a voltage control method for a pulse power supply, the voltage control method for a pulse power supply comprising:

[0006] Obtaining an input voltage of a pulse power supply, and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2;

[0007] Constructing an initial pulse frequency modulation model, and dynamically pulse controlling the N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences;

[0008] The N groups of interleaved pulse sequences are processed by a harmonic suppression algorithm to obtain pulse outputs with low harmonic distortion;

[0009] Performing high-precision voltage sampling on the low harmonic distortion pulse output to obtain real-time voltage feedback data;

[0010] Based on the real-time voltage feedback data, a multi-loop nested control strategy is used to collaboratively optimize the voltage outer loop and the current inner loop to generate a control compensation signal;

[0011] According to the control compensation signal, the parameters of the initial pulse frequency modulation model are dynamically adjusted to obtain a target pulse frequency modulation model, and the output voltage of the pulse power supply is adjusted by the target pulse frequency modulation model.

[0012] In a second aspect, the present invention provides a voltage control device for a pulse power supply, the voltage control device for the pulse power supply comprising:

[0013] An acquisition module, used for acquiring an input voltage of a pulse power supply, and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2;

[0014] A construction module, used for constructing an initial pulse frequency modulation model, and performing dynamic pulse control on the N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences;

[0015] A processing module, used for processing the N groups of interleaved pulse sequences by using a harmonic suppression algorithm to obtain a pulse output with low harmonic distortion;

[0016] A sampling module, used for performing high-precision voltage sampling on the low harmonic distortion pulse output to obtain real-time voltage feedback data;

[0017] A generation module, used to collaboratively optimize the voltage outer loop and the current inner loop based on the real-time voltage feedback data by adopting a multi-loop nested control strategy to generate a control compensation signal;

[0018] The regulating module is used to dynamically adjust the parameters of the initial pulse frequency modulation model according to the control compensation signal to obtain a target pulse frequency modulation model, and to regulate the output voltage of the pulse power supply through the target pulse frequency modulation model.

[0019] The third aspect of the present invention provides a computer device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned pulse power supply voltage control method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for controlling the voltage of a pulse power supply.

[0021] In the technical solution provided by the present invention, by performing multi-phase shunting processing on the input voltage, N-way voltage sources are obtained, which can effectively disperse the power pressure and improve the power handling capacity and efficiency of the system. At the same time, the multi-phase structure can reduce the stress of each power unit and extend the life of the system. Based on the initial pulse frequency modulation model, the N-way voltage source is dynamically pulse controlled to generate N groups of staggered pulse sequences, which can achieve more flexible output voltage regulation and improve the dynamic response capability of the system. The harmonic suppression algorithm is used to process the N groups of staggered pulse sequences, which can effectively reduce the harmonic distortion of the output voltage and improve the power quality. Real-time voltage feedback data is obtained by high-precision voltage sampling, which provides accurate input information for subsequent control and helps to improve control accuracy. The multi-loop nested control strategy is used to coordinately optimize the voltage outer loop and the current inner loop, which can achieve better voltage regulation performance and current protection function. The parameters of the initial pulse frequency modulation model are dynamically adjusted according to the control compensation signal to obtain the target pulse frequency modulation model, which can achieve adaptive control and improve the adaptability and stability of the system under different working conditions. Multi-objective optimization is achieved, while improving the output voltage accuracy, the dynamic performance and anti-interference ability of the system can also be taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0023] Figure 1 Schematic diagram of the steps of the voltage control method of the pulse power supply in an embodiment of the present invention;

[0024] Figure 2 Schematic diagram of the structure of the voltage control device of the pulse power supply in the embodiment of the present invention. DETAILED DESCRIPTION

[0025] Embodiments of the present invention provide a voltage control method, device, equipment and storage medium for a pulse power supply. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0026] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the voltage control method of the pulse power supply in the embodiment of the present invention includes:

[0027] Step S1, obtaining the input voltage of the pulse power supply, and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2;

[0028] It is understandable that the execution subject of the present invention may be a voltage control device of a pulse power supply, or may be a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0029] Specifically, the input voltage of the pulse power supply is obtained, and the input voltage is Fourier transformed to obtain the spectrum component of the input voltage. Based on the spectrum component, the input voltage is bandpass filtered to extract the required frequency components to obtain a filtered voltage signal. The filtered voltage signal is phase separated to generate N reference phase signals, and the phase interval of the reference phase signal is 2π / N to ensure the uniform distribution of the signal. According to the N reference phase signals, the filtered voltage signal is phase modulated to obtain N initial phase modulation signals. The N initial phase modulation signals are amplitude normalized to obtain N normalized phase modulation signals, so that each signal is on the same amplitude reference, which is convenient for subsequent phase and amplitude adjustment. Based on the N normalized phase modulation signals, the phase offset of each signal is calculated, and the normalized signal is phase compensated according to the calculation result to obtain a more accurate N phase compensation signal. Phase compensation can eliminate or reduce the error caused by phase offset and ensure the synchronization and accuracy of the signal. The N-channel phase compensation signals are amplified to the required voltage level to obtain N-channel amplitude modulation signals. According to the amplified N-channel amplitude modulation signals, the corresponding mathematical model is constructed to generate the actual N-channel voltage source through the digital-to-analog converter (DAC). The N-channel voltage source is used as the output of the pulse power supply, thereby realizing the precise control and regulation of the pulse power supply.

[0030] Step S2, constructing an initial pulse frequency modulation model, and dynamically controlling N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences;

[0031] Specifically, Fourier analysis is performed on N voltage sources to obtain the spectrum characteristics of each voltage source. Based on the spectrum characteristics, bandwidth limitation is performed on each voltage source to remove unnecessary high-frequency or low-frequency components, effectively reduce the noise components in the signal, and obtain a bandwidth-limited voltage source signal. Discrete sampling is performed on the bandwidth-limited voltage source signal to generate a discrete time series. The continuous signal is converted into data at a series of discrete time points so that accurate calculations can be performed during the digital processing process. Based on the discrete time series, the root mean square value of each voltage source is calculated to evaluate the effective voltage of each voltage source and determine the amplitude characteristics of the signal. According to the root mean square value, each voltage source is normalized to obtain a unified normalized voltage sequence to ensure that the signal amplitudes of all voltage sources are within the same range. The normalized voltage sequence is Hilbert transformed to generate an analytical signal. Through the Hilbert transform, the instantaneous phase and instantaneous frequency of the signal are extracted. Based on the instantaneous phase and instantaneous frequency extracted from the analytical signal, an initial pulse frequency modulation model is constructed. The form of the model is Mi(f)=A*exp(-(f-fi[n])² / (2σ²)), where Mi(f) represents the pulse frequency modulation model, f is the frequency variable, A is the amplitude coefficient, σ is the frequency modulation bandwidth parameter, and fi[n] is the instantaneous frequency. The model can accurately describe the frequency modulation characteristics of each voltage source and control the flexibility and accuracy of the modulation by adjusting the bandwidth parameter σ. Based on the initial pulse frequency modulation model, pulse width modulation (PWM) is performed on each voltage source to generate an initial PWM signal. The PWM signal controls the voltage output by adjusting the width of the pulse to achieve precise control of the power supply. In order to ensure that the phases of multiple PWM signals do not overlap, the initial PWM signal is phase-interleaved to obtain N groups of interleaved pulse sequences. The interleaved pulse sequence can effectively reduce harmonic interference and improve the power supply efficiency and stability of the system.

[0032] Step S3, using a harmonic suppression algorithm to process the N groups of interleaved pulse sequences to obtain a pulse output with low harmonic distortion;

[0033] Specifically, a fast Fourier transform is performed on N groups of interleaved pulse sequences to obtain the spectrum of each group of sequences. By analyzing the spectrum, the amplitudes of each harmonic component are calculated, and these amplitudes are important parameters for evaluating the harmonic characteristics of the pulse sequence. The amplitudes of the harmonic components directly affect the total harmonic distortion of the pulse sequence. The amplitudes of each harmonic component are weighted and summed to obtain the total harmonic distortion index (THDi), which can quantify the overall harmonic distortion of the pulse sequence. In order to effectively suppress harmonics, a harmonic suppression objective function J (α) is constructed based on the total harmonic distortion index. The expression of the objective function is J (α) = Σ (wi*THDi²), where wi is the weight coefficient of each channel, which is used to adjust the weight of the influence of different harmonic components on the total distortion, α is the harmonic suppression parameter vector to be optimized, and THDi represents the total harmonic distortion index of each group of pulse sequences. The objective function minimizes the total harmonic distortion index by adjusting the harmonic suppression parameters to achieve the optimal harmonic suppression effect. The harmonic suppression objective function is optimized by using the gradient descent algorithm. Through repeated iterative calculations, the harmonic suppression parameter α is gradually adjusted so that the objective function J (α) reaches the minimum value. In this process, each iteration will calculate the gradient based on the current parameter vector, and update the parameters along the direction of gradient descent until the optimal harmonic suppression parameters are converged. After obtaining the optimal harmonic suppression parameters, these parameters are used to construct a harmonic suppression filter. The harmonic suppression filter is applied to N groups of interleaved pulse sequences, and these sequences are filtered to eliminate or weaken the harmonic components therein to obtain the filtered pulse sequence. Based on the filtered pulse sequence, the power spectral density of each output is calculated. The power spectral density reflects the energy distribution of the pulse signal in the frequency domain and is a key indicator for evaluating the filtering effect. The power spectral density is synthesized in multiple phases to obtain the total output power spectral density. The total output power spectral density integrates the spectral characteristics of all pulse sequences and can fully reflect the spectral performance of the entire pulse power system. Based on the total output power spectral density, the frequency domain signal is reconstructed back to the time domain through inverse Fourier transform to obtain a pulse output with low harmonic distortion.

[0034] Step S4, performing high-precision voltage sampling on the pulse output with low harmonic distortion to obtain real-time voltage feedback data;

[0035] Specifically, the pulse output with low harmonic distortion is subjected to bandpass filtering to remove unnecessary frequency components and retain the main spectral components of the signal to obtain a filtered signal. Oversampling is performed based on the filtered signal to increase the time resolution of the signal by increasing the sampling rate to obtain an oversampling sequence. The oversampling sequence is quantized to convert the continuous analog signal into a discrete digital signal to obtain a quantized digital sequence. During the quantization process, the discretization of the signal will inevitably introduce quantization noise. The quantization noise power is calculated based on the quantized digital sequence to evaluate the error introduced during the quantization process. In order to improve the quality of the signal, the quantized digital sequence is subjected to dithering to generate a dithering sequence. Dithering helps to average the quantization error and reduce systematic distortion by adding appropriate random noise during the quantization process. Modulation is performed based on the dithering sequence to obtain a 1-bit quantization sequence. 1-bit quantization is a very simplified digital representation method that can effectively reduce the amount of data while retaining the main information of the signal. The 1-bit quantization sequence is subjected to digital filtering to smooth and suppress the noise of the sequence through a digital filter to obtain a high-resolution digital sequence. The high-resolution digital sequence is subjected to wavelet transform to generate a wavelet coefficient sequence. Wavelet transform is a time-frequency analysis tool that can analyze signals at different scales and extract local features and change trends of signals. Threshold denoising is performed on the wavelet coefficient sequence. By setting an appropriate threshold, the noise components contained in the coefficients are removed, and useful signal information is retained to obtain the denoised wavelet coefficient sequence. Based on the denoised wavelet coefficient sequence, an inverse wavelet transform is performed to restore the signal from the wavelet domain to the time domain, and finally the real-time voltage feedback data is obtained.

[0036] Step S5: Based on the real-time voltage feedback data, a multi-loop nested control strategy is used to collaboratively optimize the voltage outer loop and the current inner loop to generate a control compensation signal;

[0037] Specifically, the real-time voltage feedback data is digitally filtered to effectively remove high-frequency noise in the voltage signal and obtain a smoother filtered voltage signal. The voltage error signal is calculated based on the filtered voltage signal. The voltage error signal is the difference between the target voltage and the actual voltage, reflecting the current deviation of the system. The voltage error signal is PID controlled to generate a voltage outer loop control quantity. The PID controller adjusts the voltage error through three control actions: proportional, integral and differential, and generates an appropriate control quantity so that the voltage of the system gradually tends to the target value. Based on the generated voltage outer loop control quantity, the current command generation process is performed to obtain a current reference signal. The current reference signal is the target of the subsequent current control loop, which is to enable the current of the system to respond to the requirements of voltage control, thereby realizing the coordinated control of voltage and current. The actual current is sampled and digitally filtered to obtain a filtered current signal. The filtering process can eliminate noise interference in the current signal, making the current signal more accurate. Based on the current reference signal and the filtered current signal, the current error signal is calculated to reflect the deviation between the current current and the target current. The current error signal is processed by model predictive control to obtain a predicted control quantity. Model predictive control is a control algorithm that optimizes the current control quantity by predicting the future system behavior based on the dynamic model of the system, so that the system can better track the reference signal. Sliding mode control is performed based on the predicted control quantity to obtain the sliding mode control quantity. Sliding mode control is a nonlinear control strategy that enhances the robustness of the system to uncertainty and interference by maintaining sliding motion on the sliding surface of the system. The sliding mode control quantity and the predicted control quantity are weighted and fused to obtain the comprehensive control quantity. Combining the advantages of the two control methods, the accuracy of model predictive control is retained, and the robustness of sliding mode control is introduced, resulting in a more stable and reliable control effect. Pulse width modulation (PWM) is performed based on the comprehensive control quantity to generate a control compensation signal. Pulse width modulation controls the output voltage by adjusting the width of the pulse to achieve precise power supply regulation.

[0038] Step S6: dynamically adjust the parameters of the initial pulse frequency modulation model according to the control compensation signal to obtain a target pulse frequency modulation model, and adjust the output voltage of the pulse power supply by using the target pulse frequency modulation model.

[0039] Specifically, the control compensation signal is subjected to a fast Fourier transform to obtain the spectrum characteristics of the compensation signal. Through spectrum analysis, the main frequency components in the compensation signal and the amplitudes corresponding to these components are identified. The main frequency components reflect the most significant signal characteristics in the control system. The extracted main frequency components are clustered and divided into K frequency clusters. Cluster analysis can aggregate similar frequency components together to simplify the complexity of the model. Based on the frequency cluster, the initial pulse frequency modulation model and the control compensation signal are cross-correlated and the correlation coefficient between the two is obtained. The correlation coefficient measures the correlation between the compensation signal and the model parameters. Based on the correlation coefficient, the model parameters are optimized using the gradient descent method. The gradient descent method iteratively adjusts the parameters so that the objective function (such as the model error) is gradually reduced and finally converges to the optimal solution. The optimized parameter set includes the adjusted frequency, amplitude and bandwidth parameters (wk', fk', σk'), which are used to reconstruct the pulse frequency modulation model. Through the updated parameter set, the reconstructed pulse frequency modulation model can better adapt to the dynamic changes of the system and provide more accurate voltage regulation capabilities. After the reconstruction is completed, the new pulse frequency modulation model is subjected to harmonic analysis to calculate the harmonic coefficients. The purpose of harmonic analysis is to identify and quantify the harmonic components in the model, which usually have an adverse effect on the performance of the system. Based on the harmonic coefficients, a harmonic compensation function C(f) is constructed in the form of C(f)=Σ(αn*sin(2πnf / f0+φn)), where αn is the amplitude compensation parameter of the nth harmonic, f0 is the fundamental frequency, and φn is the phase compensation parameter of the nth harmonic. The function of the harmonic compensation function is to offset or weaken the harmonic components in the reconstructed model, thereby improving the output quality of the system. The reconstructed pulse frequency modulation model M1(f) is superimposed with the harmonic compensation function C(f) to obtain the target pulse frequency modulation model Mtarget(f). The expression of the target model is Mtarget(f)=M1(f)+β*C(f), where β is the mixing coefficient, which is used to adjust the proportion between the reconstructed model and the harmonic compensation function. By reasonably adjusting β, the best balance between accuracy and harmonic suppression effect can be found. The output voltage of the pulse power supply is adjusted through the target pulse frequency modulation model. The target model not only includes the reconstructed optimization parameters, but also incorporates the harmonic compensation function, which can effectively reduce the harmonic distortion in the output voltage and improve the overall performance of the pulse power supply.

[0040] In the embodiment of the present invention, by performing multi-phase shunting processing on the input voltage to obtain N-way voltage sources, the power pressure can be effectively dispersed and the power handling capacity and efficiency of the system can be improved. At the same time, the multi-phase structure can reduce the stress of each power unit and extend the life of the system. Based on the initial pulse frequency modulation model, the N-way voltage source is dynamically pulse controlled to generate N groups of staggered pulse sequences, which can achieve more flexible output voltage regulation and improve the dynamic response capability of the system. The harmonic suppression algorithm is used to process the N groups of staggered pulse sequences, which can effectively reduce the harmonic distortion of the output voltage and improve the power quality. Real-time voltage feedback data is obtained by high-precision voltage sampling, which provides accurate input information for subsequent control and helps to improve control accuracy. The multi-loop nested control strategy is used to coordinately optimize the voltage outer loop and the current inner loop, which can achieve better voltage regulation performance and current protection function. The parameters of the initial pulse frequency modulation model are dynamically adjusted according to the control compensation signal to obtain the target pulse frequency modulation model, which can achieve adaptive control and improve the adaptability and stability of the system under different working conditions. Multi-objective optimization is achieved, while improving the output voltage accuracy, it can also take into account the dynamic performance and anti-interference ability of the system.

[0041] In a specific embodiment, the process of executing step S1 may specifically include the following steps:

[0042] Acquire the input voltage of the pulse power supply, perform Fourier transform on the input voltage, obtain the frequency spectrum component of the input voltage, and perform bandpass filtering on the input voltage based on the frequency spectrum component to obtain a filtered voltage signal;

[0043] Performing phase separation processing on the filtered voltage signal to obtain N reference phase signals, wherein the phase interval of the reference phase signals is 2π / N;

[0044] According to N reference phase signals, phase modulate the filtered voltage signal to obtain N initial phase modulated signals, and perform amplitude normalization processing on the N initial phase modulated signals to obtain N normalized phase modulated signals;

[0045] Based on N normalized phase modulated signals, a phase offset of each normalized phase modulated signal is calculated, and according to the phase offset, phase compensation is performed on the N normalized phase modulated signals to obtain N phase compensated signals;

[0046] Amplitude amplification processing is performed on N-channel phase compensation signals to obtain N-channel amplitude modulation signals, and a corresponding mathematical model is constructed based on the N-channel amplitude modulation signals. According to the mathematical model, N-channel voltage sources are generated through a digital-to-analog converter.

[0047] Specifically, the input voltage signal of the pulse power supply is obtained, and the input voltage signal is Fourier transformed to convert it from the time domain to the frequency domain to obtain the spectrum components of the input voltage. The Fourier transform decomposes the voltage signal in the time domain into sine or cosine wave components of different frequencies, each component corresponding to a specific frequency and amplitude. , its Fourier transform can be expressed as:

[0048] ;

[0049] in, is the voltage signal in the frequency domain, It's time. is the frequency, is an imaginary unit. The spectrum components obtained by Fourier transform can show the amplitude and phase information of each frequency component in the voltage signal. Based on the spectrum components, the input voltage signal is bandpass filtered. The function of the bandpass filter is to allow signals within a certain frequency range to pass through, while attenuating signals of other frequencies. For example, if there is some high-frequency noise or low-frequency drift in the voltage signal, the bandpass filter can effectively remove these unnecessary frequency components and obtain a purer filtered voltage signal. The filtered voltage signal is subjected to phase separation processing to obtain The phase interval of the reference phase signal is The phases of these signals are evenly distributed between 0 and The purpose of phase separation is to decompose the filtered signal into components with different phases for subsequent phase modulation processing. The value is 4, then the phase of the reference phase signal is 0, , and .according to A reference phase signal is used to phase modulate the filtered voltage signal to obtain Phase modulation uses the reference phase signal to adjust the phase of the filtered voltage signal so that each signal has a specific phase offset. For example, if the filtered signal is a sine wave, after phase modulation, The initial phase modulation signals will be ,in It is The phase offset of the signal. The amplitude of each initial phase modulated signal is normalized to adjust the amplitude of each signal to a uniform standard range, usually between 0 and 1. Normalization helps eliminate the amplitude difference between signals and ensure the consistency of subsequent processing. Phase modulated signals are taken from the receiver and the phase offset of each signal is calculated. Phase offset refers to the degree of deviation of the signal phase relative to a certain reference phase. Calculating the phase offset helps to further adjust the phase of the signal so that the phases of all signals can be accurately aligned in a specific frequency band. Phase compensation is performed on the phase modulated signal to obtain Phase compensation signal. Phase compensation corrects the error caused by phase modulation by adjusting the phase of the signal to ensure that the phase of the signal is consistent with the expected one. The phase compensation signal is amplified and the amplitude of the signal is restored to the range of the original voltage signal. Usually the amplification factor depends on the specific needs of the system to ensure that the signal has enough energy to maintain its effectiveness in the subsequent digital-to-analog conversion and signal processing. The corresponding mathematical model is constructed by using the amplitude modulation signal of each channel. This mathematical model is used to describe the behavior characteristics of each channel signal in the time domain or frequency domain, which can usually be expressed as:

[0050] ;

[0051] in, It is road signal, is the amplitude, is the frequency, is the phase. Based on the constructed mathematical model, the digital-to-analog converter (DAC) generates The digital-to-analog converter converts the digital signal into an analog voltage signal, which is modulated and compensated. The line signal is converted into the output signal of the power supply in this process.

[0052] In a specific embodiment, the process of executing step S2 may specifically include the following steps:

[0053] Performing Fourier analysis on N voltage sources to obtain the frequency spectrum characteristics of each voltage source, and based on the frequency spectrum characteristics, performing bandwidth limitation processing on each voltage source to obtain a bandwidth-limited voltage source signal;

[0054] Discrete sampling is performed on the bandwidth-limited voltage source signal to obtain a discrete time series, and the root mean square value of each voltage source is calculated based on the discrete time series;

[0055] According to the root mean square value, each voltage source is normalized to obtain a normalized voltage sequence, and the normalized voltage sequence is Hilbert transformed to obtain an analytical signal;

[0056] Based on the analytical signal, the instantaneous phase and instantaneous frequency of each voltage source are extracted, and the initial pulse frequency modulation model is constructed according to the instantaneous frequency; the initial pulse frequency modulation model is: Mi(f)=A*exp(-(f-fi[n]) 2 / (2σ 2 )), Mi (f) is the initial pulse frequency modulation model, f is the frequency variable, A is the amplitude coefficient, σ is the frequency modulation bandwidth parameter, and fi [n] is the instantaneous frequency;

[0057] Based on the initial pulse frequency modulation model, pulse width modulation is performed on each voltage source to obtain an initial PWM signal, and phase interleaving processing is performed on the initial PWM signal to obtain N groups of interleaved pulse sequences.

[0058] Specifically, Fourier analysis is performed on N voltage sources to obtain their spectrum characteristics. Fourier analysis is a technique that converts time domain signals into frequency domain signals and can decompose complex time signals into multiple frequency components. Assume that each voltage source signal is , through Fourier transform, we can get its representation in the frequency domain ,in Indicates time, Represents frequency. The expression of Fourier transform is:

[0059] ;

[0060] in, It is The representation of the voltage source in the frequency domain is: is an imaginary unit, It is the kernel function of Fourier transform. After obtaining the spectrum characteristics of each voltage source through Fourier analysis, the main frequency components and possible noise in the signal are identified. Based on the spectrum characteristics, each voltage source is bandwidth-limited. The frequency range of the signal is limited to a specified bandwidth through a bandwidth-limiting filter, removing unnecessary high-frequency or low-frequency components and retaining useful frequency components. For example, if the effective signal frequency of a voltage source is concentrated between 50Hz and 100Hz, the bandwidth-limiting filter can attenuate other frequency components to obtain a purer bandwidth-limited voltage source signal. Discrete sampling is performed on the bandwidth-limited voltage source signal, the continuous signal is converted into a discrete time series, and the signal is converted into a digital signal form. Assume that the sampling frequency is , then the discretized signal can be expressed as ,in Represents a discrete time series index. A discrete time series obtained based on discretized sampling , calculate the root mean square (RMS) value of each voltage source, which is a measure of signal strength. The calculation formula is:

[0061] ;

[0062] in, is the number of sampling points, It is The signal value of each sampling point. Each voltage source is normalized to standardize the amplitude of each signal, usually scaling the amplitude of the signal to between 0 and 1. Standardization helps eliminate the amplitude differences between different voltage sources, making subsequent processing more unified and simple. The normalized voltage sequence can be expressed as ,in is the normalized signal value. The normalized voltage sequence is Hilbert transformed to obtain the analytical form of the signal. The Hilbert transform is a special transform that can generate a corresponding complex signal from a real signal, making the phase and envelope information of the signal more obvious. Through the Hilbert transform, the analytical signal obtained can be expressed as ,in yes The Hilbert transform of the analytic signal , extract the instantaneous phase and instantaneous frequency of each voltage source. Instantaneous phase and instantaneous frequency It is the key feature of the analytical signal. The instantaneous phase can be obtained by calculating the angle of the complex signal:

[0063] ;

[0064] The instantaneous frequency is the derivative of the instantaneous phase with respect to time:

[0065] ;

[0066] After obtaining the instantaneous frequency, the initial pulse frequency modulation model is constructed. The form of the model is:

[0067] ;

[0068] in, It is The initial pulse frequency modulation model of the line voltage source, is a frequency variable, is the amplitude coefficient, is the frequency modulation bandwidth parameter, is the instantaneous frequency. Based on the constructed initial pulse frequency modulation model, pulse width modulation (PWM) is performed on each voltage source to generate an initial PWM signal. PWM is a method of controlling the output voltage by adjusting the duty cycle of the signal. The PWM of each signal is the modulation result of the pulse frequency modulation model. The initial PWM signal is phase interleaved to ensure that the N PWM signals do not overlap in the time domain, thereby reducing harmonic interference and obtaining N groups of interleaved pulse sequences.

[0069] In a specific embodiment, the process of executing step S3 may specifically include the following steps:

[0070] Performing fast Fourier transform on N groups of interleaved pulse sequences to obtain the frequency spectrum of each group of sequences, and calculating the amplitude of each harmonic component based on the frequency spectrum of each group of sequences;

[0071] The harmonic component amplitudes are weighted and summed to obtain the total harmonic distortion index, and based on the total harmonic distortion index THDi, the harmonic suppression objective function J(α) is constructed, and the expression is: J(α)=Σ(wi*THDi 2 ), where wi is the weight coefficient of each channel, α is the harmonic suppression parameter vector to be optimized, and THDi is the total harmonic distortion index;

[0072] Perform gradient descent optimization on the harmonic suppression objective function to obtain the optimal harmonic suppression parameters, and construct a harmonic suppression filter based on the optimal harmonic suppression parameters;

[0073] Applying a harmonic suppression filter to the N groups of interleaved pulse sequences to obtain a filtered pulse sequence, and calculating the power spectral density of each output based on the filtered pulse sequence;

[0074] The power spectrum density is synthesized in multiple phases to obtain the total output power spectrum density, and the time domain signal is reconstructed through inverse Fourier transform according to the total output power spectrum density to obtain a pulse output with low harmonic distortion.

[0075] Specifically, yes Fast Fourier transform is performed on the interleaved pulse sequences to convert each pulse sequence from the time domain to the frequency domain. Through Fourier transform, the pulse sequence is decomposed into the superposition of different frequency components. For each pulse sequence , which is expressed in the frequency domain as ,in represents the index of the sequence, It's time. is the frequency. The formula for Fourier transform is:

[0076] ;

[0077] in, Indicates The spectrum of the group pulse train, is an imaginary unit, It is the kernel function of Fourier transform. By performing Fourier transform on each group of pulse sequences, the spectrum of the pulse sequence is obtained, and each harmonic component and its corresponding amplitude are extracted from it. Harmonics refer to frequency components in the signal, which are usually integer multiples of the fundamental frequency. These harmonic components often cause electromagnetic interference and power loss in power systems and electronic equipment, and need to be analyzed and suppressed. The amplitudes of the harmonic components are weighted and summed to calculate the total harmonic distortion index (THDi). The total harmonic distortion index is used to measure the degree of harmonic distortion of the entire signal and is an important indicator for measuring signal quality. Assume that the harmonic amplitude of each group of pulse sequences is , where i represents the sequence index and k represents the harmonic order, the calculation formula of the total harmonic distortion index can be expressed as:

[0078] ;

[0079] in, is the amplitude of the fundamental frequency component, It is The amplitude of the subharmonic. Based on the total harmonic distortion index THDi, the harmonic suppression objective function is constructed The objective function aims to minimize the total harmonic distortion index by optimizing the harmonic suppression parameters, thereby improving the signal quality. The expression of the objective function is:

[0080] ;

[0081] in, is the harmonic suppression objective function, is the weight coefficient of each signal, reflecting the influence of different signals on the total distortion. is the harmonic suppression parameter vector to be optimized. In order to minimize the objective function, the gradient descent method is used to Optimize. Gradient descent method is an iterative optimization algorithm that calculates the gradient of the objective function with respect to the parameter and gradually adjusts the parameter along the direction of gradient descent until the minimum value of the objective function is found. The optimal harmonic suppression parameters are obtained through gradient descent optimization. Finally, a harmonic suppression filter is constructed based on the optimal parameters. The harmonic suppression filter effectively reduces or eliminates the harmonic components in the signal by filtering the signal, thereby improving the signal quality. The filtered pulse sequence is obtained from the group of interleaved pulse sequences. The filtered sequence shows lower harmonic distortion in the frequency domain, and the signal quality is significantly improved. Based on the filtered pulse sequence, the power spectrum density of each output is calculated. The power spectrum density describes the power distribution of the signal in the frequency domain. Calculate the power spectrum density of each pulse sequence Finally, the power spectrum density is superimposed through multi-phase synthesis to obtain the total output power spectrum density of the entire system. Polyphase synthesis is a technique used to combine multiple spectral components into an overall spectrum to evaluate the frequency domain characteristics of the entire signal. Based on the total output power spectral density, the signal is reconstructed from the frequency domain back to the time domain through an inverse Fourier transform. The formula for the inverse Fourier transform is:

[0082] ;

[0083] in, is the signal representation in the frequency domain, It is the signal reconstructed in the time domain. Through inverse Fourier transform, a pulse output signal with low harmonic distortion is obtained.

[0084] In a specific embodiment, the process of executing step S4 may specifically include the following steps:

[0085] Performing bandpass filtering on the pulse output with low harmonic distortion to obtain a filtered signal, and performing oversampling processing based on the filtered signal to obtain an oversampling sequence;

[0086] Quantizing the oversampled sequence to obtain a quantized digital sequence, and calculating the quantization noise power based on the quantized digital sequence;

[0087] Performing dithering processing on the quantized digital sequence according to the quantization noise power to obtain a dithering sequence, and performing modulation processing based on the dithering sequence to obtain a 1-bit quantization sequence;

[0088] Perform digital filtering on the 1-bit quantized sequence to obtain a high-resolution digital sequence, and perform wavelet transform on the high-resolution digital sequence to obtain a wavelet coefficient sequence;

[0089] Threshold denoising is performed on the wavelet coefficient sequence to obtain a denoised wavelet coefficient sequence, and an inverse wavelet transform is performed based on the denoised wavelet coefficient sequence to obtain real-time voltage feedback data.

[0090] Specifically, the pulse output signal with low harmonic distortion is bandpass filtered. The bandpass filter allows the signal within the specified frequency range to pass through, while attenuating other unnecessary frequency components. The goal of bandpass filtering is to eliminate residual noise or unnecessary frequency components in the pulse signal to obtain a purer and more accurate filtered signal. Assume that the center frequency of the filter is And the bandwidth is , the filtered signal It can be expressed as:

[0091] ;

[0092] in, is the spectrum of the original pulse signal, is the transfer function of the bandpass filter. The filtered signal is oversampled. More signal details are captured by increasing the sampling rate, which is usually multiple times the signal bandwidth. Assume that the oversampling rate is times, the oversampled sequence can be expressed as ,in is the discrete time index of the oversampled sequence. Oversampling can increase the resolution of the signal, so that the changes in the signal can be captured more accurately in the subsequent digital processing stage. After quantization, the continuous signal value is converted into a discrete digital signal. Each sample value is mapped to the nearest discrete level to form a quantized digital sequence. However, the quantization process inevitably introduces quantization noise, the power of which is It can be calculated by the following formula:

[0093] ;

[0094] in, is the quantization interval, that is, the difference between adjacent quantization levels. Based on the quantization noise power , perform dithering on the quantized digital sequence. Dithering reduces the correlation of quantization errors by adding random noise during the quantization process, smoothes the quantization error, and obtains a dithering sequence The jittered signal usually appears smoother in visual or auditory applications, but in voltage signal processing, it is mainly used to reduce the impact of quantization error on signal accuracy. Perform modulation processing to convert the jitter sequence into a 1-bit quantization sequence This form is often used in efficient digital signal processing applications. The 1-bit quantization sequence represents the signal amplitude information by adjusting the pulse density or duty cycle. Perform digital filtering to obtain a high-resolution digital sequence . The digital filter extracts the precise signal amplitude from the 1-bit quantized signal by suppressing noise and spurious signals, thereby increasing the effective resolution of the signal. Wavelet transform processing is performed based on the high-resolution digital sequence. Wavelet transform is a signal processing technique that can decompose the signal into wavelet coefficients of different scales and analyze the characteristics of the signal at different time scales. Through wavelet transform, the wavelet coefficient sequence is obtained. ,in Indicated in Multi-scale coefficients at each time point. Threshold denoising is performed on the wavelet coefficient sequence, and the noise components below a certain threshold in the wavelet coefficient are set to zero, the effective components in the signal are retained, and the noise is removed. The wavelet coefficient sequence after threshold denoising It shows stronger signal characteristics and less noise interference. Based on the denoised wavelet coefficient sequence, the inverse wavelet transform is performed to reconstruct the time domain signal. The inverse wavelet transform restores the wavelet coefficients to the original signal form to obtain the final real-time voltage feedback data. .

[0095] In a specific embodiment, the process of executing step S5 may specifically include the following steps:

[0096] Performing digital filtering on the real-time voltage feedback data to obtain a filtered voltage signal, and calculating a voltage error signal based on the filtered voltage signal;

[0097] Perform PID control processing on the voltage error signal to obtain the voltage outer loop control quantity, and perform current command generation processing based on the voltage outer loop control quantity to obtain the current reference signal;

[0098] Sampling and digitally filtering the actual current to obtain a filtered current signal, and calculating a current error signal based on the current reference signal and the filtered current signal;

[0099] Performing model predictive control on the current error signal to obtain a predicted control quantity, and performing sliding mode control processing based on the predicted control quantity to obtain a sliding mode control quantity;

[0100] The sliding mode control quantity and the predictive control quantity are weightedly fused to obtain a comprehensive control quantity, and pulse width modulation is performed based on the comprehensive control quantity to obtain a control compensation signal.

[0101] Specifically, the real-time voltage feedback data is digitally filtered to remove noise and unstable components in the feedback data to obtain a smoother and more reliable voltage signal. Through a digital filter, such as a low-pass filter, the high-frequency noise components are attenuated, leaving only the low-frequency components in the voltage signal. The voltage error signal is calculated based on the filtered voltage signal. The voltage error signal is the expected voltage The actual measured voltage The difference between them is calculated as:

[0102] ;

[0103] Voltage error signal It reflects the voltage deviation of the system and is the core variable of voltage control. The voltage error signal is processed by PID control. The PID controller adjusts the voltage error through three control methods: proportional, integral and derivative, to generate the voltage outer loop control quantity. The control formula of the PID controller can be expressed as:

[0104] ;

[0105] in, is the proportional gain, is the integral gain, is the differential gain. By adjusting these three parameters, the PID controller can balance the response speed, stability and accuracy of the system. Based on the voltage outer loop control quantity, the current command generation process is performed to convert the voltage control into current control. Based on the voltage outer loop control quantity, the current reference signal is calculated This signal determines the target current required by the system. After sampling and digital filtering the actual current, the filtered current signal is obtained. The role of digital filtering is similar to that of voltage signal processing, which aims to remove noise from the current signal and ensure the stability and accuracy of the current signal. And the filtered current signal , calculate the current error signal The calculation formula of the current error signal is:

[0106] ;

[0107] Current error signal It reflects the deviation between the actual current and the target current, which requires further control adjustment. In order to better control the current, the current error signal is subjected to model predictive control (MPC). Model predictive control generates an optimized control strategy by predicting the system behavior at future moments, so that the system can better track the reference signal. MPC uses the mathematical model of the system to predict the system output for a period of time in the future based on the current state, and obtains the predicted control quantity through the optimization algorithm. The model predictive control can provide forward-looking regulation for the system, making the system more stable and efficient in dynamic response. Perform sliding mode control processing to obtain the sliding mode control quantity Sliding mode control is a nonlinear control method that can effectively deal with the uncertainty and external interference in the system by controlling the system to slide on the sliding surface. Usually the sliding surface The sliding surface can be defined as a linear combination of the current error and its derivative:

[0108] ;

[0109] in, is the design parameter of the sliding mode control, which is used to adjust the influence of the sliding surface on the system response. and predictive control Perform weighted fusion processing to obtain the comprehensive control quantity The calculation formula of comprehensive control quantity is:

[0110] ;

[0111] in, is the weighting coefficient, which determines the proportion of predictive control and sliding mode control in the comprehensive control quantity. Perform pulse width modulation (PWM) to obtain control compensation signal PWM controls the average voltage of the output signal by adjusting the width of the pulse to achieve precise control of the load. The duty cycle of the PWM signal is determined by the comprehensive control quantity The decision ensures that the output current and voltage can accurately track the reference signal to achieve stable and efficient operation of the system.

[0112] In a specific embodiment, the process of executing step S6 may specifically include the following steps:

[0113] Performing fast Fourier transform on the control compensation signal to obtain the spectrum characteristics of the compensation signal, and extracting the main frequency components and corresponding amplitudes based on the spectrum characteristics of the compensation signal;

[0114] Perform cluster analysis on the main frequency components to obtain K frequency clusters, and based on the frequency clusters, perform cross-correlation analysis on the initial pulse frequency modulation model and the control compensation signal to obtain the correlation coefficient;

[0115] Based on the correlation coefficient, the model parameters are optimized by using the gradient descent method to obtain an updated parameter set, and the pulse frequency modulation model is reconstructed according to the updated parameter set to obtain a reconstructed pulse frequency modulation model;

[0116] Performing harmonic analysis on the reconstructed pulse frequency modulation model to obtain harmonic coefficients, and constructing a harmonic compensation function based on the harmonic coefficients;

[0117] The reconstructed pulse frequency modulation model and the harmonic compensation function are superimposed to obtain the target pulse frequency modulation model, and the output voltage of the pulse power supply is adjusted by the target pulse frequency modulation model, wherein the target pulse frequency modulation model is: Mtarget(f)=M1(f)+β*C(f), M1(f) is the reconstructed pulse frequency modulation model, expressed as: M1(f)=Σ(wk'*exp(-(f-fk') 2 / (2σk' 2 )), wk', fk', σk' are optimized parameters, C(f) is the harmonic compensation function, expressed as: C(f)=Σ(αn*sin(2πnf / f0+φn)), where: n represents the harmonic order, αn is the amplitude compensation parameter of the nth harmonic, f0 is the fundamental frequency, φn is the phase compensation parameter of the nth harmonic, and β is the mixing coefficient, which is used to adjust the proportion of the reconstructed pulse frequency modulation model and the harmonic compensation function.

[0118] Specifically, the control compensation signal is subjected to a fast Fourier transform to convert the signal from the time domain to the frequency domain. The Fourier transform can decompose the complex signal in the time domain into a series of sine and cosine wave components of different frequencies, each of which corresponds to a specific frequency and amplitude. Through the Fourier transform, the spectral characteristics of the compensation signal are obtained. ,in represents the frequency, Represents the amplitude of the signal at this frequency. By analyzing the spectrum characteristics , extract the main frequency components and their corresponding amplitudes. Perform cluster analysis on the extracted main frequency components. Cluster analysis is a method of grouping data, classifying frequency components with similar frequency characteristics into the same category. Assume that the main frequency components are divided into frequency clusters, each of which corresponds to a group of similar frequency components. Cluster analysis helps the system identify the frequency characteristics in the signal, so that the subsequent modulation model can more accurately match the actual spectrum distribution of the signal. Frequency clusters are generated to perform cross-correlation analysis on the initial pulse frequency modulation model and the control compensation signal. Cross-correlation analysis is used to measure the similarity or correlation between two signals and helps determine the degree of match between the modulation model and the actual signal at different frequencies. Through cross-correlation analysis, the correlation coefficient is calculated. , which reflects the frequency The larger the correlation coefficient, the better the modulation effect of the model at this frequency. , the parameters of the modulation model are optimized using the gradient descent method. The gradient descent method is a commonly used optimization algorithm that calculates the gradient of the objective function with respect to the parameters and adjusts the parameters along the direction of gradient descent to gradually approach the minimum value of the objective function. The objective function can be defined as the error measure between the modulation model and the signal. By iteratively updating the parameters, the optimized parameter set is finally obtained. ,in is the weight parameter, is the center frequency, is the frequency bandwidth. Based on the updated parameter set, the pulse frequency modulation model is reconstructed The expression of this model is:

[0119] ;

[0120] in, It is a reconstructed pulse frequency modulation model, which describes the amplitude distribution of the modulated signal at different frequencies. Harmonic analysis is performed on the pulse frequency modulation model to identify and quantify the harmonic components in the model, that is, those frequency components that are integer multiples of the fundamental frequency. Harmonic components often introduce additional distortion and noise. Through harmonic analysis, the harmonic coefficients are obtained. , describing the amplitude and phase characteristics of each harmonic. Based on the harmonic coefficients Constructing harmonic compensation function , the harmonic compensation function is used to correct the harmonic distortion in the modulation model, and its expression is:

[0121] ;

[0122] in, Indicates the harmonic order, is the fundamental frequency, It is The phase compensation parameters of subharmonics. The reconstructed pulse frequency modulation model With harmonic compensation function Superposition is performed to obtain the target pulse frequency modulation model The expression of the target model is:

[0123] ;

[0124] in, is the mixing coefficient, which is used to adjust the proportion of the reconstructed model and the harmonic compensation function in the target model. , and adjust the output voltage of the pulse power supply. This model combines optimized frequency modulation and harmonic compensation, which can effectively reduce harmonic distortion and improve the stability and accuracy of the power supply output.

[0125] The voltage control method of the pulse power supply in the embodiment of the present invention is described above. The voltage control device of the pulse power supply in the embodiment of the present invention is described below. Figure 2 , an embodiment of the voltage control device of the pulse power supply in the embodiment of the present invention includes:

[0126] An acquisition module is used to acquire the input voltage of the pulse power supply and perform multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2;

[0127] A construction module is used to construct an initial pulse frequency modulation model, and dynamically control the pulses of N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences;

[0128] A processing module, used for processing N groups of interleaved pulse sequences by using a harmonic suppression algorithm to obtain pulse outputs with low harmonic distortion;

[0129] The sampling module is used to perform high-precision voltage sampling on the pulse output with low harmonic distortion to obtain real-time voltage feedback data;

[0130] A generation module is used to collaboratively optimize the voltage outer loop and the current inner loop based on real-time voltage feedback data using a multi-loop nested control strategy to generate a control compensation signal;

[0131] The regulating module is used to dynamically adjust the parameters of the initial pulse frequency modulation model according to the control compensation signal to obtain the target pulse frequency modulation model, and to regulate the output voltage of the pulse power supply through the target pulse frequency modulation model.

[0132] Through the cooperation of the above components, the input voltage is processed in multiple phases to obtain N voltage sources, which can effectively disperse the power pressure and improve the power handling capacity and efficiency of the system. At the same time, the multi-phase structure can reduce the stress of each power unit and extend the life of the system. Based on the initial pulse frequency modulation model, the N voltage sources are dynamically pulse controlled to generate N groups of staggered pulse sequences, which can achieve more flexible output voltage regulation and improve the dynamic response capability of the system. The harmonic suppression algorithm is used to process the N groups of staggered pulse sequences, which can effectively reduce the harmonic distortion of the output voltage and improve the power quality. Real-time voltage feedback data is obtained through high-precision voltage sampling to provide accurate input information for subsequent control, which helps to improve the control accuracy. The multi-loop nested control strategy is used to coordinately optimize the voltage outer loop and the current inner loop, which can achieve better voltage regulation performance and current protection function. The parameters of the initial pulse frequency modulation model are dynamically adjusted according to the control compensation signal to obtain the target pulse frequency modulation model, which can realize adaptive control and improve the adaptability and stability of the system under different working conditions. Multi-objective optimization is achieved, which can improve the output voltage accuracy while taking into account the dynamic performance and anti-interference ability of the system.

[0133] The present invention also provides a computer device, which includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor executes the steps of the voltage control method of the pulse power supply in the above-mentioned embodiments.

[0134] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the voltage control method of the pulse power supply.

[0135] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0136] If the integrated unit is implemented in the form of 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 the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0137] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A voltage control method for a pulse power supply, characterized in that: The voltage control method of the pulse power supply comprises: Obtaining an input voltage of a pulse power supply, and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2; Construct an initial pulse frequency modulation model, and perform dynamic pulse control on the N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences; specifically, the method comprises: performing Fourier analysis on the N voltage sources to obtain the spectrum characteristics of each voltage source, and based on the spectrum characteristics, performing bandwidth limitation processing on each voltage source to obtain a bandwidth-limited voltage source signal; performing discrete sampling on the bandwidth-limited voltage source signal to obtain a discrete time series, and calculating the root mean square value of each voltage source based on the discrete time series; performing normalization processing on each voltage source according to the root mean square value to obtain a normalized voltage series, and performing Hilbert transform on the normalized voltage series to obtain an analytical signal; based on the analytical signal, extracting the instantaneous phase and instantaneous frequency of each voltage source, and constructing an initial pulse frequency modulation model according to the instantaneous frequency; the initial pulse frequency modulation model is: Mi(f)=A*exp(-(f-fi[n]) 2 / (2σ 2 )), Mi (f) is an initial pulse frequency modulation model, f is a frequency variable, A is an amplitude coefficient, σ is a frequency modulation bandwidth parameter, and fi [n] is an instantaneous frequency; based on the initial pulse frequency modulation model, pulse width modulation is performed on each voltage source to obtain an initial PWM signal, and the initial PWM signal is subjected to phase interleaving processing to obtain N groups of interleaved pulse sequences; The N groups of interleaved pulse sequences are processed by a harmonic suppression algorithm to obtain pulse outputs with low harmonic distortion; Performing high-precision voltage sampling on the low harmonic distortion pulse output to obtain real-time voltage feedback data; Based on the real-time voltage feedback data, a multi-loop nested control strategy is used to collaboratively optimize the voltage outer loop and the current inner loop to generate a control compensation signal; According to the control compensation signal, the parameters of the initial pulse frequency modulation model are dynamically adjusted to obtain the target pulse frequency modulation model, and the output voltage of the pulse power supply is adjusted by the target pulse frequency modulation model; specifically including: performing fast Fourier transform on the control compensation signal to obtain the spectrum characteristics of the compensation signal, and extracting the main frequency components and corresponding amplitudes based on the spectrum characteristics of the compensation signal; performing cluster analysis on the main frequency components to obtain K frequency clusters, and based on the frequency clusters, performing cross-correlation analysis on the initial pulse frequency modulation model and the control compensation signal to obtain a correlation coefficient; based on the correlation coefficient, using the gradient descent method to optimize the model parameters to obtain an updated parameter set, The pulse frequency modulation model is reconstructed according to the updated parameter set to obtain a reconstructed pulse frequency modulation model; harmonic analysis is performed on the reconstructed pulse frequency modulation model to obtain harmonic coefficients, and a harmonic compensation function is constructed based on the harmonic coefficients; the reconstructed pulse frequency modulation model and the harmonic compensation function are superimposed to obtain a target pulse frequency modulation model, and the output voltage of the pulse power supply is adjusted by the target pulse frequency modulation model, wherein the target pulse frequency modulation model is: Mtarget (f) = M1 (f) + β * C (f), M1 (f) is the reconstructed pulse frequency modulation model, expressed as: M1 (f) = Σ (wk' * exp (- (f-fk') 2 / (2σk' 2 )), wk', fk', σk' are optimized parameters, C(f) is the harmonic compensation function, expressed as: C(f)=Σ(αn*sin(2πnf / f0+φn)), where: n represents the harmonic order, αn is the amplitude compensation parameter of the nth harmonic, f0 is the fundamental frequency, φn is the phase compensation parameter of the nth harmonic, and β is the mixing coefficient, which is used to adjust the proportion of the reconstructed pulse frequency modulation model and the harmonic compensation function.

2. The voltage control method of the pulse power supply according to claim 1, characterized in that: The step of obtaining the input voltage of the pulse power supply and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, wherein N is an integer greater than or equal to 2, includes: Acquire an input voltage of a pulse power supply, perform Fourier transform on the input voltage to obtain a frequency spectrum component of the input voltage, and perform bandpass filtering on the input voltage based on the frequency spectrum component to obtain a filtered voltage signal; Performing phase separation processing on the filtered voltage signal to obtain N reference phase signals, wherein the phase interval of the reference phase signals is 2π / N; According to the N reference phase signals, phase modulate the filtered voltage signal to obtain N initial phase modulated signals, and perform amplitude normalization processing on the N initial phase modulated signals to obtain N normalized phase modulated signals; Based on the N normalized phase modulated signals, a phase offset of each normalized phase modulated signal is calculated, and according to the phase offset, phase compensation is performed on the N normalized phase modulated signals to obtain N phase compensated signals; The N-channel phase compensation signals are amplitude-amplified to obtain N-channel amplitude modulation signals, and corresponding mathematical models are constructed based on the N-channel amplitude modulation signals. According to the mathematical model, N-channel voltage sources are generated through a digital-to-analog converter.

3. The voltage control method of the pulse power supply according to claim 1, characterized in that: The method of using a harmonic suppression algorithm to process the N groups of interleaved pulse sequences to obtain a pulse output with low harmonic distortion includes: Performing fast Fourier transform on N groups of interleaved pulse sequences to obtain the frequency spectrum of each group of sequences, and calculating the amplitude of each harmonic component based on the frequency spectrum of each group of sequences; The harmonic component amplitudes are weighted and summed to obtain the total harmonic distortion index, and based on the total harmonic distortion index THDi, a harmonic suppression objective function J(α) is constructed, and the expression is: J(α)=Σ(wi*THDi 2 ), where wi is the weight coefficient of each channel, α is the harmonic suppression parameter vector to be optimized, and THDi is the total harmonic distortion index; Performing gradient descent optimization on the harmonic suppression objective function to obtain optimal harmonic suppression parameters, and constructing a harmonic suppression filter according to the optimal harmonic suppression parameters; Applying the harmonic suppression filter to N groups of interleaved pulse sequences to obtain filtered pulse sequences, and calculating the power spectral density of each output based on the filtered pulse sequences; The power spectrum density is subjected to multi-phase synthesis to obtain a total output power spectrum density, and a time domain signal is reconstructed through inverse Fourier transform according to the total output power spectrum density to obtain a pulse output with low harmonic distortion.

4. The voltage control method of the pulse power supply according to claim 1, characterized in that: The high-precision voltage sampling of the low harmonic distortion pulse output to obtain real-time voltage feedback data includes: Performing bandpass filtering on the pulse output with low harmonic distortion to obtain a filtered signal, and performing oversampling processing based on the filtered signal to obtain an oversampling sequence; Quantizing the oversampled sequence to obtain a quantized digital sequence, and calculating quantization noise power based on the quantized digital sequence; Performing dithering processing on the quantized digital sequence according to the quantization noise power to obtain a dithering sequence, and performing modulation processing based on the dithering sequence to obtain a 1-bit quantization sequence; Performing digital filtering processing on the 1-bit quantization sequence to obtain a high-resolution digital sequence, and performing wavelet transform processing based on the high-resolution digital sequence to obtain a wavelet coefficient sequence; Threshold denoising is performed on the wavelet coefficient sequence to obtain a denoised wavelet coefficient sequence, and an inverse wavelet transform is performed based on the denoised wavelet coefficient sequence to obtain real-time voltage feedback data.

5. The voltage control method of the pulse power supply according to claim 1, characterized in that: Based on the real-time voltage feedback data, a multi-loop nested control strategy is used to collaboratively optimize the voltage outer loop and the current inner loop to generate a control compensation signal, including: Performing digital filtering on the real-time voltage feedback data to obtain a filtered voltage signal, and calculating a voltage error signal based on the filtered voltage signal; Performing PID control processing on the voltage error signal to obtain a voltage outer loop control quantity, and performing current command generation processing based on the voltage outer loop control quantity to obtain a current reference signal; Sampling and digitally filtering the actual current to obtain a filtered current signal, and calculating a current error signal based on the current reference signal and the filtered current signal; Performing model predictive control on the current error signal to obtain a predicted control amount, and performing sliding mode control processing based on the predicted control amount to obtain a sliding mode control amount; The sliding mode control amount and the predicted control amount are subjected to weighted fusion processing to obtain a comprehensive control amount, and pulse width modulation is performed based on the comprehensive control amount to obtain a control compensation signal.

6. A voltage control device for a pulse power supply, characterized in that: A method for controlling a voltage of a pulse power supply according to any one of claims 1 to 5, wherein the voltage control device of the pulse power supply comprises: An acquisition module, used for acquiring an input voltage of a pulse power supply, and performing multi-phase shunting processing on the input voltage to obtain N voltage sources, where N is an integer greater than or equal to 2; A construction module, used for constructing an initial pulse frequency modulation model, and performing dynamic pulse control on the N voltage sources based on the initial pulse frequency modulation model to generate N groups of staggered pulse sequences; A processing module, used for processing the N groups of interleaved pulse sequences by using a harmonic suppression algorithm to obtain a pulse output with low harmonic distortion; A sampling module, used for performing high-precision voltage sampling on the pulse output of the low harmonic distortion to obtain real-time voltage feedback data; A generation module, used to collaboratively optimize the voltage outer loop and the current inner loop based on the real-time voltage feedback data by adopting a multi-loop nested control strategy to generate a control compensation signal; The regulating module is used to dynamically adjust the parameters of the initial pulse frequency modulation model according to the control compensation signal to obtain a target pulse frequency modulation model, and to regulate the output voltage of the pulse power supply through the target pulse frequency modulation model.

7. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and is characterized in that when the processor executes the computer program, the voltage control method of the pulse power supply described in any one of claims 1 to 5 is implemented. 8 . A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the processor is caused to execute the voltage control method of a pulse power supply according to any one of claims 1 to 5.

Images