96 results about "Differential algorithm" patented technology

The invention discloses a multi-objective optimization method and system for dual active bridge circuit expansion phase shift modulation, and the method comprises the steps: calculating a voltage matching ratio, and determining an external phase shift angle ratio based on a difference value between an output voltage and an output voltage reference value; according to the voltage matching ratio and the external phase shift angle ratio, obtaining an internal phase shift angle ratio of the primary side full bridge through an offline lookup table; calculating the maximum transmission power of the dual-active bridge converter, and determining a current working mode from the six preset working modes based on the size relationship between the inner phase shift angle ratio and the outer phase shift angle ratio; generating an optimal modulation strategy through a multi-objective optimization algorithm by taking the current effective value and the current peak value as optimization objectives and combining the voltage matching ratio and the periodic parameters of the current working mode; and generating a trigger signal of each switch tube of the dual-active bridge converter. Through six-mode working area division, complete differential algorithm boundary constraint and off-line-on-line collaborative optimization, the contradiction between the current stress and the soft switching range of the converter under extended phase shift modulation is solved.

The invention discloses a dam termite image recognition method based on multi-modal fusion and dynamic feature enhancement. The method comprises the steps that image data, environmental parameters and vibration signals of a dam monitoring area are collected in real time through a multi-modal sensor network; carrying out the preprocessing of the image data, and carrying out the target detection through an improved ACP-YOLOv5s model; processing the continuous video frames by adopting Gaussian background modeling and a difference algorithm to generate a binary mask graph representing dynamic activities of termites, and separating a moving target area from a static background; and the target detection result and the dynamic mask graph are fused, the collected environment parameters and vibration signals are combined, the termite damage level is calculated through an AI engine, and graded early warning or automatic killing response is triggered. The method breaks through the limitation of traditional single-mode detection. And a multi-dimensional hazard grade evaluation model is established, so that the automation level of dam termite prevention and control is greatly improved.

The invention discloses a regulation and control system and method based on heat energy of an energy storage cabinet, and belongs to the technical field of intelligent heat management. Temperature data in the energy storage cabinet are collected, a temperature distribution matrix is constructed, and a thermal gradient vector and an overheating risk area are extracted based on a difference algorithm; further establishing a thermal power load mapping model, and dynamically generating an air circulation path optimization diagram; the opening angle of a flow deflector and the rotating speed of a fan are adjusted in combination with an optimized path, and self-adaptive regulation and control of a thermal field in the cabinet body are achieved; meanwhile, a machine learning model is trained by utilizing historical operation data, an optimal regulation and control strategy is output, and an execution layer is driven to realize collaborative response of a fan and a flow deflector; the method has the advantages of high regulation and control precision, high response speed, low energy consumption and the like, and is suitable for heat management requirements of various types of energy storage cabinets.

The present invention relates to the fields of PID control and Kalman filtering, and disclosed is an improved PID control method based on Kalman filtering. The present invention comprises: in the process of acquiring a target value of an expected signal, taking the difference between the target value and a current feedback quantity of a controlled object to obtain an error value, acquiring an output of PID from the error value, applying the output of PID to the controlled object to obtain a noisy output signal, then, after Kalman filtering, feeding the output back to PID, continuously iterating to make the feedback quantity approach the target value, and finally obtaining a stable output signal. Variable-speed integration and integral separation algorithms are added to an integral term of PID to change an integration speed on the basis of a system deviation, and an incomplete differential algorithm and a low-pass filter are added to a differential term of PID to prevent high-frequency interference. According to the present invention, the optimal estimation of a system state is obtained by means of prediction of a current state and updating of observation data, thereby reducing the influence of measurement errors and noise, and improving the performance and stability of a PID controller.

The invention relates to the technical field of debugging systems, in particular to an adaptive debugging system and method for touch screen FW parameters, and the system comprises a signal acquisition module, a trend triggering module, a temperature compensation calculation module, a window generation module and a parameter output module. According to the invention, the edge of the driving voltage is carefully collected, the sliding difference algorithm is introduced to filter out high-frequency noise, meanwhile, real-time sampling of environment temperature and baseline capacitance adjustment are combined, the influence of external variables on signals can be dynamically corrected, and accurate judgment of capacitance fluctuation is improved by utilizing periodic delay deviation and regression trend analysis. By means of main frequency feature extraction and zero crossing point sudden change detection, threshold setting can be refined, self-adaptive window expansion can be achieved, touch signal response is kept sensitive and consistent all the time, adjustment and optimization can still be automatically completed even under the condition that the temperature changes frequently or signal interference is complex, parameter drift and debugging failure are effectively prevented, and user experience is improved. And the parameter automatic matching efficiency and the finished product consistency in batch production are improved.

The invention relates to the technical field of video recognition and analysis, and discloses a video recognition and analysis method based on physical characteristics. Video frame pixels are mapped to a two-dimensional coordinate system with the upper left corner as an original point, and mirror image expansion and median filtering are carried out on a gray level image; constructing a binary image based on a gray threshold value, and analyzing and extracting a target region by using a four-neighborhood connected domain; using neighborhood search and polar angle sorting to close the tracking contour, and generating equidistant re-sampling points based on Euclidean distance and an interpolation method; the curvature of the re-sampling points is estimated through a three-point difference algorithm, and zero denominator is avoided through numerical protection; performing discrete Fourier transform on the curvature sequence to extract a frequency spectrum, and normalizing an amplitude to form a standardized feature vector; and finally, inter-frame similarity is calculated based on the feature vector, and the most similar frame is automatically retrieved. By processing unified data standards in stages, edge noise is suppressed, sampling uniformity is ensured, and feature stability and cross-frame comparability are improved.

The invention relates to the technical field of metal composite materials, in particular to a metal composite material production process dynamic simulation method based on digital twinning. Comprising the following steps: constructing a digital twin model comprising a physical entity layer, a virtual model layer and a connection layer, and establishing a Fourier neural operator model embedded with a Navier Stokes equation and a heat conduction equation constraint; real-time molten pool monitoring data are mapped into model input through a connecting layer, and advanced time step deduction is executed by using a deep learning model to predict the state of a molten pool; based on the differentiable characteristic of a physical information neural operator, calculating the gradient of a prediction state and an ideal target by using an automatic differential algorithm, generating a process parameter field through reverse iteration updating, and discretizing the process parameter field into a galvanometer control instruction; and when the risk of aggregation of the reinforced phase particles is predicted to exist, automatically generating a swing scanning strategy to realize active suppression. The method gives consideration to simulation precision and real-time performance.

The invention discloses a mining area equipment increment upgrading method and system based on a difference algorithm, and the method comprises the steps: obtaining the GNSS positioning and base station fingerprint positioning of mining area equipment, carrying out the comparison verification with a preset electronic fence polygon region after preprocessing, and activating an upgrading task after the verification is passed; carrying out difference comparison on the old file and the new file by adopting an improved binary system difference algorithm through dynamic sliding window partitioning and a multi-level hash index, and compressing a difference part to generate an increment difference upgrade package; and establishing connection with mining area equipment through a network, transmitting the increment difference upgrade package to the mining area equipment by adopting a three-way handshake transmission protocol, and downloading, decompressing and upgrading the increment difference upgrade package by the mining area equipment. According to the method, directional upgrading can be realized in a complex network environment, the flow consumption can be remarkably reduced, the retransmission data volume is reduced, and the upgrading success rate of weak signal areas such as strip mine pits is improved.

A calculation method for a transformer winding considering a frequency-dependent loss, comprising: according to simulation requirements, defining key parameters calculated by using a finite-difference time-domain (FDTD) algorithm; constructing an equivalent flat wire model in an FDTD grid; calculating frequency-dependent internal impedance per unit length of a flat wire; and on the basis of a vector fitting method and a time-domain convolution calculation method, substituting the frequency-dependent internal impedance into a time-domain equation of the FDTD algorithm for updating and iterative solving, so as to obtain an electromagnetic field calculation result. In the present invention, on the basis of the vector fitting method and time-domain convolution calculation, the impact of the frequency-dependent characteristics of an electromagnetic field distributed in a conductor on an electromagnetic transient process is accurately considered during the FDTD time-domain solving, so that the calculation accuracy and efficiency of conventional FDTD algorithms and even conventional cross-scale modeling techniques can be significantly improved, thereby carrying out the targeted electromagnetic transient analysis on the reliability, safety, and predictability of transformers, and achieving the early detection of potential faults and defects, or formulating targeted maintenance plans and repair schemes.

The invention relates to the technical field of industrial data management. The data management method, system and device based on the industrial digital base and the medium are provided, the method comprises the steps that block chain evidence storage is conducted on version hash, micro-differential packet hash and an obtained operator certificate, an on-chain evidence storage block is generated, distributed storage processing is conducted on a micro-differential packet, and an off-chain differential packet storage entity is obtained; when it is detected that the data is abnormal, abnormal version positioning is carried out based on the record-level blood relationship chain, and an abnormal version sequence is determined; on the basis of a predefined inverse operation mapping relation, performing reverse playback on a micro-differential packet corresponding to the abnormal version sequence to realize historical data state recovery; and carrying out signature verification on the incremental service pack, carrying out dynamic loading processing after the verification is passed, updating a preset blood relationship extraction rule and a preset difference algorithm, and generating a hot update strategy, so as to achieve the technical effects of improving the data blood relationship tracing precision, reducing the storage redundancy overhead and enhancing the credibility of an evolution process.

The invention provides an intelligent HAST equipment adjusting method and system based on multi-source heterogeneous data and a medium, the heterogeneous data is collected in real time based on temperature, humidity and pressure sensors, the instantaneous change rate and the sliding window mean value are calculated through a difference algorithm, the optimal cooling mode is matched in combination with a preset rule base, and the optimal cooling mode is obtained. The problems of control lag and misoperation caused by the fact that traditional equipment depends on a single threshold value are solved; in the initialization stage, a modeling driver parameter template is loaded, dynamic adjustment is conducted according to the real-time change rate in the operation process, the fan rotating speed is optimized through a fuzzy PID algorithm, an ARIMA time sequence model is introduced to predict the future trend, the exhausting or cooling action is triggered in advance, and passive response is avoided; and when the parameter mean value exceeds a safety threshold value, executing emergency shutdown in a grading manner and recording an over-limit event to ensure the safety of equipment and samples.

The invention provides a process GPU-oriented real-time fault tolerance and task migration system and method, and relates to the technical field of distributed computation.The method comprises the steps that a training task state in a GPU video memory is captured, difference data is generated through an incremental difference algorithm, and the difference data is stored in a multi-level storage system after being subjected to lightweight compression; meanwhile, GPU node health indexes are collected in real time, node states are analyzed through a time sequence prediction model, and a migration process is triggered when fault risks are predicted or preemption early warning is received; during migration, the task is paused, a final increment check point is generated, check point data is quickly transmitted to a target healthy GPU node through an RDMA network, a complete context state is reconstructed, task execution is recovered, and real-time fault tolerance and seamless migration of the GPU task are achieved. According to the method, the continuous operation reliability and the resource utilization efficiency of the GPU cluster in the long-time training task are improved.

The present invention discloses a method for improving the prediction accuracy of soil iron oxide based on an amplified hyperspectral dataset. The method comprises the following steps: amplifying the hyperspectral dataset and performing multiple fractional-order differential transformations on the original hyperspectral spectrum using a fractional-order differential algorithm; then combining the fractional-order differential spectra with the original spectrum to form different amplified hyperspectral databases to amplify the amount of original spectral data; establishing a 1D-CNN neural network model for each of the different amplified hyperspectral databases, where the independent variable is the hyperspectral reflectance value and the dependent variable is the soil iron oxide content; and selecting the optimal 1D-CNN model based on the models established for the different amplified databases. The present invention effectively solves the problem of low accuracy of deep learning models under small sample conditions, improves the applicability and effectiveness of deep learning models to a certain extent, effectively improves the existing prediction accuracy of soil iron oxide based on hyperspectral data, and provides new methods and ideas for hyperspectral prediction research of other soil properties.

The invention provides a power plant DCS data acquisition interface automatic configuration method and system. The method comprises the following steps: acquiring a production process diagram file from a DCS upper computer, and analyzing and extracting measuring point information based on a DCS type; generating a structured lexicographical order through lexical analysis, and screening target measurement points according to a preset rule to generate a measurement point list; adopting a Myers difference algorithm to compare the states of the measuring point lists, and dynamically updating the interface program acquisition range; and transmitting the measuring point list, the process diagram and the real-time data to a management information layer MIS across a safety area, and synchronously checking configuration consistency based on a digital twinborn model. According to the method, automatic extraction of the measuring points is realized through a regular expression and a DOM analysis technology, dynamic key encryption and an incremental updating mechanism are combined, the problems of low manual configuration efficiency, updating lag and cross-regional data islands are solved, the accuracy and synchronous real-time performance of measuring point configuration are remarkably improved, and the method is suitable for large-scale popularization and application. And safe and reliable digital support is provided for power plant production monitoring and information management.

The invention relates to the technical field of optical measurement, in particular to an optical scattering method based on an inverse problem solving algorithm and a mixed measurement method for batch samples thereof, and the method comprises the following steps: carrying out optical measurement on samples to obtain actually measured optical signals; obtaining prior structure information of the sample; constructing an initial model; analyzing the initial model by utilizing strict coupling wave analysis to generate a simulation optical signal; judging whether the difference between the generated simulation optical signal and the actually measured optical signal meets a set requirement or not; if yes, outputting a measurement result; and if not, gradient calculation is carried out on each parameter in the initial model, gradients of all the parameters are obtained by utilizing an automatic differential algorithm, then the initial model is optimized by utilizing the gradients of the parameters to obtain an optimized model, and the step of analyzing by utilizing the strict coupling waves is repeatedly executed until a set requirement is met. According to the method, more detailed contour optimization can be provided, and the calculation efficiency can be improved.

The invention discloses a parallel robot trajectory optimization method based on an improved dream optimization algorithm. The invention relates to the technical field of parallel robot trajectory optimization, and aims to obtain a smooth running trajectory of a robot in a high-speed running process and improve the motion performance of the parallel robot in the high-speed running process. The method comprises the following steps: constructing a target function according to total time and total impact required when a parallel robot operates according to a specified track; constructing constraint conditions according to the speed and acceleration operation requirements of the parallel robot; initializing a population of a dream optimization algorithm by means of a good point set and reverse learning; improving a dream sharing strategy by adopting a difference algorithm; and jumping out of the local optimization capability through the crossover variation improvement algorithm, thereby improving the quality of the solution and obtaining the optimal solution of the target function. The objective function considers the running time and impact of the robot, and the stability of the running process of the robot is guaranteed while the running speed of the robot is increased.

The application discloses a kind of data mirror persistent backup systems of power grid, it is related to electric power system technical field, including data acquisition module, data preprocessing module, difference data synchronization module, data integrity check module, data mirror module, data storage module, control and management module;Data acquisition module is used to collect the operation data of power grid in real time;Data preprocessing module carries out cleaning, formatting and compression processing to the data collected, prepares for subsequent data synchronization;Difference data synchronization module calculates and identifies the change part of data since last synchronization by difference algorithm, and only synchronizes the change part data.The application greatly reduces the data transmission amount when real-time topology reconstruction by only synchronizing the changed data, reduces the backup time at the same time, improves the efficiency of power grid data backup, optimizes the synchronization process of data, reduces the storage space waste caused by full backup, and guarantees the accuracy of backup data by hash value comparison mode.

The invention discloses a nuclear pulse signal trapezoidal forming method suitable for an FPGA, and the method comprises the steps: carrying out the high-speed digital sampling of an exponential decay original signal outputted by a detector, and obtaining a discrete digital detector signal vn; calculating the discrete digital detector signal vn by using a differential algorithm to obtain a differentiated pulse signal pn; delaying the pulse signal pn by using a digital delay line to form three delayed pulse signals, and performing synthesis operation on the pulse signal pn and the delayed pulse signals to obtain a synthesized pulse signal sequence sn; and performing integral operation on the synthesized pulse sequence signal to obtain a trapezoidal forming pulse yn. The nuclear pulse signal trapezoid forming method is simple in calculation and suitable for being implemented in the FPGA, few FPGA logic resources are used, and the calculation speed is high.

The present invention discloses an active filter protection method, comprising: obtaining the current flowing through the capacitor branch of the filter under normal circumstances; obtaining the action value of the double-ended differential protection based on the capacitor branch current; and determining whether to activate the differential protection based on the relationship between the action value and the action range. The present invention can protect the entire filter, eliminating dead zones in the protection range and operating conditions. Simultaneously, the use of a two-side differential algorithm provides stronger braking for out-of-zone faults and higher sensitivity for in-zone faults. Compared with multi-side differential protection, this method significantly improves the sensitivity and reliability of the protection, reduces the computational complexity and difficulty of setting coordination, and has wide applicability and high practicality.

The application is a target detection and parameter estimation method of a sensing integrated system, aiming to solve the target detection and fractional Doppler and time delay parameter estimation problems of an OTFS sensing integrated system, and the method comprises the following steps: step one, a sensing integrated signal is generated at the sending end; step two, the receiver receives the echo signal and processes it; step three, a threshold map is generated through an integer time delay and Doppler detection network based on a convolutional neural network to perform target detection; and step four, fractional Doppler and time delay indexes are estimated. The target detection and parameter estimation method of the OTFS sensing integrated system based on deep learning has the following advantages: compared with the commonly used CFAR detection algorithm, the detection performance of the integer time delay and Doppler detection network is better. In addition, compared with the differential algorithm considering only integer Doppler and time delay and considering fractional Doppler and time delay, the relative distance and speed estimated by the fractional estimation network are more accurate.

The application provides a power plant DCS data acquisition interface automatic configuration method and system. The method comprises the following steps: obtaining a production process diagram file from a DCS host computer, analyzing and extracting measuring point information based on a DCS type; generating a structured dictionary sequence through lexical analysis, and generating a measuring point list according to a preset rule to screen target measuring points; comparing the measuring point list state by using a Myers difference algorithm to dynamically update the interface program acquisition range; transmitting the measuring point list, the process diagram and real-time data to a management information layer (MIS) across a safety zone, and synchronously checking the consistency based on a digital twin model. The application realizes automatic extraction of measuring points by using regular expressions and DOM analysis technology, solves the problems of low efficiency of manual configuration, lagging update and cross-zone data island by combining dynamic key encryption and incremental update mechanism, significantly improves the accuracy of measuring point configuration and synchronization real-time performance, and provides safe and reliable digital support for power plant production monitoring and information management.

The application belongs to the technical field of computers, and provides a PDF document comparison method and system based on intelligent differential merging. The method is as follows: the visual fingerprint sequences of a source PDF document and a target PDF document are calculated; the visual fingerprint sequences of the two PDF documents are compared through a Myers differential algorithm, a difference sequence is generated, all deletion operation sequence regions adjacent to insertion operation sequence regions are identified, and a candidate page pair is constructed; pre-filtering is performed on the candidate page pair, a similarity score is calculated, and based on a threshold judgment and a greedy strategy, the best match is locked, the deletion operation node and the insertion operation node are combined and replaced by a modification operation node; finally, spatial quantization index technology is used to compare internal objects, and the final comparison result is output. The application can not only accurately identify the subtle changes of the PDF document, but also reduce the comparison complexity, improve the comparison efficiency, and meet the real-time comparison needs of users.

The invention relates to a differential cloud platform and a vehicle software differential upgrading method, and the platform comprises a first storage unit, a differential engine service, a differential algorithm service and a second storage unit. The first storage unit is used for storing a difference algorithm mirror image file; the second storage unit is used for storing a vehicle software package file; the difference engine service is used for obtaining the difference algorithm mirror image file from the first storage unit in response to an input difference request so as to start the difference algorithm service based on the difference algorithm mirror image file; and the differential algorithm service is used for obtaining the vehicle software package file from the second storage unit and carrying out differential processing on the vehicle software package file through a differential algorithm program to generate a differential software package, so that the problem that the adaptation difference of different differential software architectures in platform integration cannot be shielded is solved.

This invention discloses a guidance and control method for high-dynamic aircraft based on a fusion differential method. To improve the system's response speed and continuous accuracy, this method designs a fusion differential algorithm that includes nonlinear and linear terms to ensure rapid response in the initial response stage and smooth tracking in the subsequent tracking stage during line-of-sight angular velocity estimation. This allows for timely and accurate acquisition of the line-of-sight angular velocity required for guidance and control. Furthermore, based on a gravity-compensated proportional guidance method, the method designs the required overload for the high-dynamic aircraft. The required overload derives control commands according to the transfer ratio of different aircraft and transmits them to the servo mechanism. The servo mechanism then controls the servo motor to control the aircraft's flight, achieving higher-precision guidance and control.

This disclosure relates to the field of oil and gas exploration technology, and in particular to a method, apparatus, equipment, medium, and product for suppressing anomalous noise. The method includes: performing effective signal inversion on seismic data based on the L1 norm to obtain an initial expression for the inverted effective signal; and using a mathematical differential algorithm to obtain a regularization term for calculating the effective signal; confirming the expression for the inverted effective signal based on the initial expression and the regularization term; and solving for the inverted effective signal using the Lagrange multiplier method based on the expression for the inverted effective signal to remove anomalous noise and obtain the effective signal of the seismic data. By performing effective signal inversion on seismic data using the L1 norm, anomalous noise in the seismic data can be accurately identified and removed, while minimizing damage to the effective signal, effectively improving the signal-to-noise ratio of the seismic data, and avoiding damage to the effective signal due to excessive denoising.

The invention provides a slope change identification method and system based on image processing, and relates to the technical field of data processing, and the method comprises the steps: obtaining a to-be-identified slope original image; preprocessing the original image of the slope; according to the preprocessed image, generating a corresponding image sequence through frame extraction processing; extracting a moving slope image in the image sequence through a background difference algorithm; measuring the displacement of the surface of the moving slope image through an LSPIV algorithm to obtain a displacement vector field; eliminating an error vector in the displacement vector field by using a vector filtering mode; calculating an actual displacement value of a missing position in the displacement vector field through a neighborhood interpolation method; and carrying out binary segmentation on the displacement vector field, and identifying and marking the range contour of the moving slope. According to the invention, cracks, foreign matters and the like in the slope image can be sensitively identified, and identification and monitoring of a complete change area of the slope are realized.

The invention belongs to the technical field of gesture somatosensory interaction, and particularly relates to a gesture somatosensory interaction method based on fused vision, inertial touch and multi-point touch information, edge detection and a three-frame difference algorithm are combined, so that an obtained image has good contour information, and a part of'cavity 'phenomenon is made up; and meanwhile, the improved three-frame difference method is supplemented by fusing the Gaussian mixture background difference algorithm, so that the target image has continuous contour and relatively complete internal information, the gesture recognition rate is improved, and the false detection rate is reduced.

The invention discloses a three-column horizontal rotation type disconnecting switch opening and closing state monitoring method and device, and the method comprises the steps: S1, scanning a disconnecting switch working place through a laser radar, and obtaining the point cloud data of a disconnecting switch; s2, designing a point cloud difference algorithm based on an octree to process the disconnecting switch point cloud data to obtain difference four-dimensional point cloud data; s3, designing an isolating switch point cloud discrimination strategy based on semantic segmentation, and processing the differential four-dimensional point cloud data to obtain semantic enhanced four-dimensional point cloud data; s4, generating a kinematic skeleton of the conductive arm of the disconnecting switch, and resolving the motion trend of the disconnecting switch; and S5, determining the opening and closing state of the isolation switch. By adopting the technical scheme of the invention, the rapid and accurate fitting of the kinematic skeleton of the conductive arm is realized, and then the monitoring and judgment of the opening and closing states of the three-column horizontal rotary disconnecting switch are completed.

The application belongs to the technical field of industrial automation program control and digital logic control system, and relates to a logic control system for instantaneous flow of a continuous fluid conveying node, comprising: a signal acquisition module for acquiring a discrete flow signal sequence, a trend deduction module for extracting flow evolution characteristics based on a difference algorithm and generating a leading prediction flow parameter through Taylor series truncation operation in combination with a constant term representing pipe network lag; and an adjustment rhythm solving module for solving a parameter hedging instruction according to a deviation amount of the leading prediction flow parameter and a target value. The application hedges physical phase lag induced by fluid medium movement inertia through a leading prediction mechanism, eliminates time sequence misplacement of adjustment action and flow state evolution, cooperates with a logic degradation operator for discrete variance fluctuation to suppress high-frequency noise amplification effect, and ensures closed-loop adjustment stability of the conveying node under nonlinear complex disturbance working conditions.

The invention discloses a method and a system for optimizing a wide working range of a neutral-point-clamped dual-active bridge converter, and the method comprises the steps: dynamically selecting six working modes based on a voltage matching ratio and actual power, covering a full range, reducing the current stress compared with the conventional single / dual phase-shift control, and adaptively matching a boost / buck scene; taking a current effective value and a peak value as double targets, introducing a complete differential algorithm to compulsively constrain and synchronously reduce loss, expanding a soft switching range at the same time, and combining a genetic algorithm to perform global search to avoid local optimum and improve the optimization speed; the optimal inner shift ratio of different voltage matching ratios and outer shift ratios is pre-calculated and stored in an off-line table, the modulation parameter calculation time is compressed to be within 10 microseconds, a high-frequency switch is supported, the problem of online optimization divergence during voltage sudden change is solved, the primary ANPC zero state is configured through a DVDM strategy, inductive current is forced to flow through an anti-parallel diode, and the optimal inner shift ratio is stored in an off-line table. And clamping switch ZVS switching-on and non-clamping switch zero-current switching-off are realized, full-load ZVS is realized, and the method is suitable for high-dynamic scenes.