179results about How to "Improve estimation performance" patented technology

The invention discloses a pilot optimization method for large-scale MIMO channel estimation based on structural compressed sensing. The method comprises the steps of establishing a channel estimation model for a large-scale MIMO-OFDM (Multiple-Input-Multiple-Output-Orthogonal Frequency Division Multiplexing) system when pilots are placed in an overlapping mode; simplifying the channel estimation model for the large-scale MIMO-OFDM system, thereby enabling the channel estimation model to correspond to a structural compressed sensing model; and obtaining an optimum pilot matrix through utilization of a pilot optimization algorithm. Through adoption of the optimum pilot matrix, according to the channel estimation of the large-scale MIMO system based on structural compressed sensing, the mean square errors MSEs of the channel estimation are clearly reduced, and the channel estimation performance is improved.

The invention relates to a Gaussian Newton method-based amplitude-phase error and direction-of-arrival joint estimation method, which comprises the following steps of: constructing a likelihood function about array element amplitude error, array element phase error and direction-of-arrival according to an array observation model, and calculating a cost function according to the likelihood function, converting a parameter estimation problem of the likelihood function into a parameter optimization problem to obtain a target function; calculating an initial value of a direction of arrival in the target function by adopting a MUSIC algorithm, and setting an initial value of an array element amplitude error and an initial value of an array element phase error; and according to the initial value of the direction of arrival, the initial value of the array element amplitude error and the initial value of the array element phase error, jointly estimating the array element amplitude error, the array element phase error and the direction of arrival by adopting a Gauss-Newton method to obtain an estimated value of the array element amplitude error, an estimated value of the array element phase error and an estimated value of the direction of arrival. According to the method, the influence of the array amplitude-phase error on the DOA estimation precision is reduced, and the DOA estimation performance is improved.

The invention discloses a space-time near field source three-dimensional parameter estimation method based on cumulant domain step-by-step decoupling, and the method comprises the steps: receiving a signal through a cross array, and carrying out the phase shift through Fresnel approximation; the method comprises the following steps: constructing a four-order cumulant matrix by considering signal amplitude attenuation and combining with signals received by a cross array, and decomposing a three-dimensional parameter estimation problem into a one-dimensional parameter estimation problem through step-by-step decoupling: firstly, decomposing a pitch angle and an azimuth angle of an incident signal into included angles with an x axis and a y axis; then angle spectrum peak searchers constructed by a reduced-rank MUSIC algorithm are adopted for estimation, then distance spectrum peak searchers are constructed for distance estimation, and finally the pitch angle and the azimuth angle are calculated according to the estimated values of the included angles between the incident signals and the x axis and the y axis, so that three-dimensional parameter estimation is achieved. The method has the advantages that the influence of signal amplitude attenuation is considered, the high-order cumulant is used, three-dimensional parameter estimation is achieved, and the estimation precision is higher.

The invention discloses an MIMO system channel estimation method and device based on ODDM modulation, and the method comprises the steps: generating a pilot frequency data signal, carrying out the ODDM modulation of the pilot frequency signal, and then transmitting the modulated signal to a receiving module through a transmitting module. The receiving module performs matched filtering on the received pilot signal to obtain a received signal; an equivalent input and output relation is derived according to a generated pilot signal and a channel received by a receiving end, and a low-complexity approximate Bayesian optimal algorithm is provided to obtain information such as time delay, Doppler, gain and angle of the channel. According to the method, the accuracy of channel estimation can be improved in a high-speed moving scene, and meanwhile, the process of matrix inversion in channel estimation is also avoided, so that the calculation complexity is remarkably reduced.

The invention discloses a three-dimensional space PM2.5 concentration estimation method and device based on an improved UNet, and the method comprises the following steps: employing an unmanned plane to collect and obtain a haze image, employing a haze sensor to collect a PM2.5 concentration value as a data set, and carrying out the data preprocessing; and inputting the training data set into the improved UNet model for iterative training until a model iterative training loss function value is minimum, training to obtain a network model based on the improved UNet, inputting the haze image into the trained network model, and outputting a PM2.5 concentration value. According to the method, information extracted by the improved UNet network is fused, fine-grained haze reconstruction in a three-dimensional space can be realized, and the method has better generalization performance.

The invention discloses a system Gaussian filter design method for multi-step measurement of random delay and colored measurement of noise. A first-order autoregression model is introduced to whiten the colored measurement noise, and a system with the colored measurement noise is converted into a nonlinear system with normal white noise; modeling delay measurement data by means of a Bernoulli random variable, processing multi-step random delay by taking new system measurement noise as a state increment, realizing one-step prediction of delay measurement information, and finally establishing a nonlinear system state model and a measurement model with multi-step measurement random delay and colored measurement noise. And an unscented transformation mode is adopted to approximate non-Gaussian weighted integral, and an unscented Kalman filtering algorithm under the conditions of multi-step measurement random delay and colored measurement noise is provided. According to the method, the accuracy of the filter in processing multi-step measurement random delay and colored measurement noise is improved, the deviation in a filtering result is reduced, and higher precision and better numerical stability are achieved.

The invention belongs to the technical field of retired battery sorting, and particularly relates to a retired battery secondary sorting method, system and equipment based on random charging fragments and a medium. The method comprises the steps of generating a single battery identification code, constructing a battery code database, and performing charging and discharging control on a single battery based on a preset charging and discharging strategy. And obtaining a time sequence sample data set of temperature, voltage and capacity change, converting the time sequence sample data set into two-dimensional image data and storing the two-dimensional image data, and training a prediction model through a two-dimensional convolutional neural network to estimate the maximum available capacity of the battery so as to realize secondary sorting of the battery. The system comprises a code scanning module, a charging and discharging control module, a data acquisition module, a data processing module and a sorting module. The method can reduce test time and energy consumption, improve battery sorting efficiency and precision, reduce memory occupation and rear-end model operation time, and is suitable for efficient and accurate sorting of retired batteries.

The invention relates to an adaptive sparse channel estimation method applied to a reconfigurable intelligent surface (RIS)-assisted millimeter wave large-scale MIMO system. According to the method, two progressive algorithms, namely an NLMS-LOG algorithm and an NLMS-NLMF-LOG algorithm, are designed, and the method aims at making full use of the sparsity of millimeter wave channels so as to improve the accuracy and convergence speed of channel estimation. According to the NLMS-LOG algorithm, a logarithm and a penalty term are innovatively introduced into a cost function, so that the convergence speed is effectively increased; the NLMS-NLMF-LOG algorithm fuses the fast convergence advantage of the NLMS algorithm and the excellent performance of the NLMF algorithm under the condition of low signal-to-noise ratio, and the precision of channel estimation is further improved. Simulation experiment results prove that the scheme provided by the invention is obviously superior to the prior art in the aspects of channel estimation precision and convergence speed.

The invention relates to the technical field of motor control, in particular to an integral self-adaptive observer method for permanent magnet synchronous motor sensorless control, which comprises the following steps of: 1, taking a surface-mounted permanent magnet synchronous motor as a research object, selecting a two-phase static coordinate system under a two-phase static coordinate system, the error i between the actual value and the estimated value of the stator current is a state variable; 2, designing a finite time integral sliding mode function s, and enabling the system to converge in finite time; a smooth H (i) function is designed, the function weakens buffeting caused by constant high switching gain, a finite time integral sliding mode function is further provided, a fuzzy rule is designed to adjust back electromotive force feedback gain, the sine degree of back electromotive force is made to be better, the estimation performance of the rotor position and speed is improved, and the method is suitable for large-scale popularization and application. And the accurate observation performance is still kept under the complex working condition.

The invention relates to a scene-proportion-unknown super-large-scale MIMO mixed field channel estimation method based on an OMP algorithm. The method comprises the following steps: constructing a signal model from a sending end to a receiving end; calculating a Rayleigh distance according to the transmitting end antenna array and the receiving end antenna array; constructing a far-field channel model and a near-field channel model between the transmitting end and the receiving end according to the calculated Rayleigh distance; constructing a mixed field channel model between the transmitting end and the receiving end according to the constructed far-field channel model and near-field channel model; and solving the mixed field channel model by using an OMP algorithm with an unknown scene proportion, and estimating the mixed field channel between the sending end and the receiving end. According to the method, the signal can be effectively estimated in a scene in which the proportion of far-field and near-field path numbers is unknown.

The invention relates to a method for estimating high-resolution atmosphere carbon dioxide concentration in a land area of China, which comprises the following steps of: (1) calculating a vegetation index, a surface water index and a vegetation phenology index by using satellite data, and generating a vegetation data file; (2) generating a carbon dioxide background field file and a meteorological background field file by using a CAMS global greenhouse gas concentration data set and a fifth-generation atmosphere reanalysis data set of the European mid-term weather forecast center; (3) generating a human-derived carbon dioxide emission data file by using a multi-scale emission list of the Xinghua University; (4) constructing an optimized atmospheric carbon dioxide simulation system; and (5) inputting the vegetation data file, the carbon dioxide background field and meteorological background field file and the man-made source carbon dioxide emission data file into the optimized atmospheric carbon dioxide simulation system, and operating to obtain a high-resolution atmospheric carbon dioxide concentration estimation result in the land area of China. The method can significantly improve the ability to estimate the carbon dioxide concentration of the atmosphere in the land area of China.

The invention belongs to the field of ballistic target tracking, and discloses an interrupted track association method and system based on combination of a dynamic model and pseudo measurement, which can more accurately estimate the target state even under the condition of data missing or batch skipping by combining the dynamic model and the pseudo measurement. According to the method, the abnormal condition in the measurement data can be effectively processed, the track interruption risk caused by data missing is reduced, and the stability of the system is improved. According to the method, pseudo measurement of polynomial function fitting is utilized, the original track can be updated more effectively, errors are reduced, and continuity and consistency of the track are improved. The method is suitable for track association in various targets and complex environments, and the association strategy can be dynamically adjusted according to actual conditions. According to the method, the track fuzzy similarity is calculated and the association cost matrix is constructed, so that a more scientific basis can be provided for track matching, and an association result is optimized.

The invention discloses an information physical system distributed security state estimation method and device and a storage medium, and belongs to the technical field of information physical systems, and the information physical system distributed security state estimation method comprises the following steps: carrying out observable decomposition on a system dynamic model to obtain a local observable matrix pair; obtaining a gain matrix of the distributed observer according to the local observable matrix pair and the local unobservable matrix; performing communication based on the local system state vector to obtain a local system state vector after communication; obtaining an output system state vector of the distributed observer based on the gain matrix and the local system state vector after communication; and controlling an adaptive attack compensation function in the distributed observer to compensate the spoofing attack signal so as to realize that an output system state vector is consistent with a system state vector in the system dynamic model. According to the method, the problem of state estimation of the information physical system deployed on a large scale under spoofing attack interference can be solved.

The invention provides a five-phase FTFSCW-IPM novel approach rate FTSMC control method based on time-varying ESO, and the method comprises the following steps: building a five-phase FTFSCW-IPM motor mathematical model which comprises a flux linkage, a voltage, a torque and a motion equation; designing a fast terminal sliding mode surface and an improved reaching law to obtain a control law of a five-phase FTFSCW-IPM improved reaching rate fast terminal sliding mode in a rotating speed loop, and performing controller stability analysis in combination with a Lyapunov theorem; a time-varying parameter ESO is designed to be used for five-phase FTFSCW-IPM disturbance observation estimation, the characteristics and parameters of an observer are analyzed, an improved index reaching rate control method is introduced, a feedforward compensation strategy based on a time-varying extended state observer is combined, and the robustness and anti-interference performance of a five-phase FTFSCW-IPM motor control system are improved.

The invention discloses a rotating speed control method of a permanent magnet synchronous motor. The method comprises the following steps: firstly, designing an adaptive switching expansion state observer of the permanent magnet synchronous motor, and adding the adaptive switching expansion state observer into a rotating speed loop control process; and then inputting the rotating speed of the permanent magnet synchronous motor and a q-axis current reference value output by a rotating speed ring into an extended state observer, performing adaptive parameter switching control at the same time, and outputting the rotating speed and an observed value of lumped disturbance after processing so as to continue to perform rotating speed ring control and realize rotating speed control of the permanent magnet synchronous motor. The method can improve the dynamic and steady state performance of the permanent magnet synchronous motor under inertia mismatch, weaken the overshoot of the motor in the dynamic process, greatly accelerate the response speed, reduce the rotation speed fluctuation and oscillation in the adjustment process, reduce the dependence of the parameter design of a motor control system on the rotational inertia parameter, and improve the stability of the motor control system. And meanwhile, the estimation capability of rapid change lumped disturbance is enhanced, and more complex lumped disturbance types can be dealt with.

The present invention provides a channel estimation method for a MIMO communication system assisted by an intelligent reflecting surface, applies deep learning technology to the channel estimation of the intelligent reflecting surface, and introduces a conditional generative adversarial network with a received signal as conditional information, thereby effectively improving the accuracy of channel estimation and improving robustness performance by reducing information loss during neural network training; adopts a U‑Net structure and a patching structure in the generator and the discriminator respectively to improve the ability of the generator to generate more realistic data and the ability to identify input detailed information; proposes an optimization loss function for updating the parameters of the generator to maintain the correct optimization direction during the training process, thereby further improving the channel estimation performance of the conditional generative adversarial network in the adversarial training process of the generator and the discriminator. The channel estimation method for a MIMO communication system assisted by an intelligent reflecting surface based on the conditional generative adversarial network of the present invention effectively improves the channel estimation performance of the MIMO communication system assisted by the intelligent reflecting surface.

The invention relates to the technical field of aircraft system fault detection, in particular to an aircraft system fault detection method based on an unknown input observer. According to the method, joint estimation of the state variable and the fault signal of the nonlinear aircraft system can be realized, and the influence of external disturbance on the estimation performance is effectively reduced. According to the method, the defects of an existing fault detection method based on residual signals can be overcome, the fault occurrence time can be provided, the estimated value of the fault amplitude can also be provided, the fault can be further processed in the later period, and the method has wider and profound theoretical significance and practical application value; estimated values of state variables and fault signals of the system can be provided at the same time, faults can be detected in time, and a certain foundation can be laid for monitoring dynamic behaviors of the system in real time and completing specific control tasks.

The invention relates to an STCA-MIMO radar unambiguous parameter estimation method based on dimension reduction MUSIC. The method comprises the following steps: acquiring an echo signal at a target by using an STCA-MIMO radar signal emission model; obtaining a corresponding noise subspace according to the echo signal; obtaining an angle estimation matrix Q1 according to the receiving steering vector of the radar signal emission model at any point in the space; constructing a spatial spectrum function according to the angle estimation matrix Q1, and performing one-dimensional angle search on the spatial spectrum function to obtain a target angle estimation value of the angle where the target is located; constructing a spatial spectrum estimation matrix Qv2 according to the target angle estimation value and the noise subspace; obtaining a transmitting end distance steering vector according to the spatial spectrum estimation matrix Qv2; and averaging the distance estimation values obtained after the phase of the distance steering vector of the transmitting end is obtained to obtain the distance estimation value of the target. According to the technical scheme, the calculation amount and time consumption are reduced, and the parameter estimation performance is improved.

The application discloses a self-disturbance rejection control method and system based on an error compensation type extended state observer, and the method comprises the following steps: a linear self-disturbance rejection controller acquires input signals and each output state of an extended state observer, and outputs a first control signal; the linear self-disturbance rejection controller acquires input signals and each output state of the extended state observer, and outputs the first control signal; an estimation error model of the extended state observer for output displacement is constructed, a controlled object is converted into an integrator series type form, and an output displacement signal is obtained based on the controlled object; the output displacement signal and a control input signal are input into the extended state observer, and each output state is obtained; and then the linear self-disturbance rejection controller and the output displacement model are fed back. The estimation capability of the observer is greatly improved without increasing the bandwidth of the observer; the error compensation type disturbance observer is combined with the self-disturbance rejection control, and the tracking performance of the self-disturbance rejection control is further improved.

The invention relates to an FDA-MIMO radar target parameter estimation method based on a mobile nested array, and belongs to the field of radar signal processing. The method comprises the following implementation steps: constructing a mobile nested array model of a monostatic FDA-MIMO radar, and carrying out angle and distance decoupling; a sample covariance matrix of an observation matrix is calculated, an optimization problem based on low-rank matrix reconstruction is established, a Toeplitz matrix is solved and obtained by using an alternating projection method, and a target angle and distance estimation value is obtained by using a two-dimensional subspace method. The method has the advantages that the mobile nested array is applied to FDA-MIMO, the number of virtual array elements of a differential common array is remarkably increased, the array aperture is larger, the target parameter estimation precision is higher by solving the low-rank matrix reconstruction problem, and the problems that an existing FDA-MIMO radar is low in parameter estimation precision and resolution, limited in detectable target number and the like can be solved.

The invention discloses a sky wave large-scale MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) channel smooth beam structure precoding transmission method and system. A sky wave large-scale MIMO-OFDM base station designs a beam structure precoder of each user in combination with a group of subcarriers on the basis of constraining the equivalent channel smoothness of each user receiving end by using a beam-based channel model, and pre-codes signals sent to each user by using the channel smooth beam structure precoder to realize downlink transmission. The method comprises the following steps of: firstly, pre-coding a data signal and a pilot signal which are sent to each user by a base station on each subcarrier by using a beam domain precoder, and mapping a beam domain pre-coded signal of each user to a complete beam domain sending signal based on a space beam set of each user; and finally, converting the sum of the beam domain sending signals of the users into a spatial domain sending signal. According to the method, the equivalent channel estimation performance can be effectively improved and the signal detection performance can be remarkably improved while the sky wave large-scale MIMO-OFDM downlink precoding transmission performance is ensured.

The invention relates to the technical field of signal processing, in particular to a multi-path channel pseudo code sequence estimation method and system based on RAKE receiving and MLSE estimation, and the method comprises the steps: carrying out the signal processing of a received signal under a multi-path channel, obtaining pseudo code period estimation, and carrying out the rough estimation of the time delay and amplitude of each path of the received signal; constructing a multi-path channel receiving signal model based on the time delay and amplitude coarse estimation of each path, so that the multi-path channel receiving signal model performs signal symbol detection and pseudo code estimation by using a self-interference cancellation (SIC) and RAKE receiving detection algorithm; an information symbol and pseudo code sequence joint estimation cost function is set based on a multi-channel receiving signal model, and iterative optimization estimation is carried out on the multi-channel receiving signal model by using maximum likelihood estimation (MLSE) based on the cost function. And taking the pseudo code sequence and the information symbol sequence corresponding to the iteration result with the minimum cost function value during model convergence as final estimation results and outputting the final estimation results. The method can be suitable for signal processing requirements under the conditions of low signal-to-noise ratio and various parameters.

The application discloses a kind of robust direction of arrival estimation method based on ADMM-Net, by expanding ADMM algorithm into model-driven deep network ADMM-Net to improve DOA estimation accuracy, speed up DOA estimation speed and have robustness to array disturbance.Firstly, the sparse transformation of source and array received multi-shot data is carried out by space over-complete dictionary, and the DOA estimation is converted into compressed sensing sparse recovery problem;Then, ADMM algorithm is expanded, and model-driven deep network ADMM-Net with explainability is formed;ADMM-Net is used to reconstruct source power spectrum and carry out DOA estimation.The present application can learn the hyperparameters in iterative algorithm and array disturbance, solve the problem that existing compressed sensing DOA estimation method based on calculation speed is slow, and fails under the condition of array disturbance, realize fast, accurate, explainable and robust DOA estimation.

The present invention discloses a broadband millimeter wave velocity measurement radar device and a high-precision direction-finding method, which are mainly used for high-precision measurement of river flow data. The device comprises a transmitting millimeter wave array antenna, a receiving millimeter wave array antenna, a transmitting filter, a power amplifier, a phase shifter, a transmitting mixer, a signal transmitter, a receiving filter, a receiving amplifier, an orthogonal mixer, an analog-to-digital converter, a signal processor, network transmission and control and other key components. The high-precision direction-finding method of the present invention is particularly targeted at the broadband signal of the millimeter wave array, and realizes high-precision direction of arrival (DOA) estimation. The method converts the DOA estimation problem into a sparse reconstruction problem by constructing a dictionary matrix of different frequency points and adopting sparse reconstruction technology, and solves it by an iterative method, thereby effectively improving the accuracy of azimuth estimation. The technical solution of the present invention not only improves the accuracy of velocity measurement, but also realizes all-round velocity measurement, and has broad application prospects.

The application discloses a deep learning DOA estimation method based on original IQ data. The application uses I and Q components of the original signal as the input of the model to improve the performance. The application aims to solve the DOA estimation problem of a single signal source, and models the single signal source DOA estimation problem as a single-label multi-classification problem. By discretizing the DOA range, the possible directions of arrival are taken as corresponding labels. A convolutional neural network is designed to adapt to different numbers of snapshots, and accurate DOA estimation can be adaptively obtained for input signals of different lengths. Experimental results show that, compared with existing deep learning DOA estimation methods based on covariance matrix as input, the scheme has more excellent performance, and can provide a more reliable solution for array signal processing.

The invention discloses a DOA estimation method based on an expanded and augmented nested array, relates to the technical field of target positioning, and lays a foundation for improving DOA estimation performance by expanding an array aperture with the help of the expanded and augmented nested array. Meanwhile, vectorization and deduplication processing are performed on a covariance matrix of a received signal, discrete Fourier transform is performed on the received signal to obtain a discrete Fourier spectrum, a discrete Fourier series is determined through global coarse search and local fine search, a DOA accurate estimation value is finally obtained, the problem of fuzzy estimation caused by array element spacing extension is avoided, and the accuracy of DOA estimation is improved. The method is more suitable for target positioning.

The invention discloses a method for assisting early judgment of adolescent idiopathic scoliosis based on a walking video. The method comprises the steps of obtaining two-dimensional key point coordinates of an AIS patient; deriving three-dimensional key point coordinates; calculating a corresponding power spectrum index; and obtaining the disease degree of the AIS patient. The method has the advantages that the AIS patient can conveniently monitor the development condition of the own illness state in real time; an AlphaPose system is adopted, and meanwhile, a DLT method is introduced, so that the capability of estimating coordinates of three-dimensional key points can be enhanced; the MAnE algorithm and the MGE algorithm are adopted, the walking characteristics of the AIS patient can be described from the two dimensions of time and space, and the characteristics of abnormal gaits can be obtained more clearly; a computer technology is applied to data acquisition, mathematics is applied to feature calculation, a doctor can be effectively assisted in AIS early diagnosis, a corresponding treatment scheme is formulated, and therefore the monitoring effect and safety can be improved, and the life quality of an AIS patient can be improved.

The present invention provides a sensorless auto-disturbance rejection control method for a permanent magnet synchronous motor with nonlinear flux linkage observation, which solves the problem of poor control effect of the permanent magnet synchronous motor, including: step 1, measuring the induced current of the permanent magnet synchronous motor to obtain a rotor position estimation value and a speed estimation value; step 2, using the speed estimation value and a given speed command value, obtaining a q-axis current command value through a speed loop auto-disturbance rejection controller; step 3, transforming the induced current to obtain a d-axis current and a q-axis current, and then obtaining a d-axis voltage command value and a q-axis voltage command value; step 4, converting the d-axis voltage command value and the q-axis voltage command value into an axis voltage command value and an axis voltage command value, and inputting them into an SVPWM modulator to control the permanent magnet synchronous motor. The present invention uses a nonlinear flux linkage observer as a position observer, uses a rotor position estimation value and a speed estimation value to calculate the rotor position and speed estimation, and realizes sensorless control of the permanent magnet synchronous motor.

The present invention discloses a distributed safety state estimation method, device and storage medium for an information-physical system, belonging to the technical field of information-physical systems. The distributed safety state estimation method for an information-physical system includes: performing observable decomposition on a system dynamic model to obtain a local observable matrix pair; obtaining a gain matrix of a distributed observer according to the local observable matrix pair and the local unobservable matrix; communicating based on a local system state vector to obtain a local system state vector after communication; obtaining an output system state vector of a distributed observer based on the gain matrix and the local system state vector after communication; and controlling an adaptive attack compensation function in the distributed observer to compensate for a deception attack signal, so as to achieve consistency between the output system state vector and the system state vector in the system dynamic model. The present invention can solve the problem of state estimation of information-physical systems deployed on a large scale under the interference of deception attacks.

The invention provides a high-robustness state estimation method for a vehicle-mounted proton exchange membrane fuel cell, and the method sequentially comprises the steps: building a fuel cell air supply system mechanism model, analyzing the sensitivity of system response to key parameters, achieving the parameter screening, constructing a multi-innovation volume-based Kalman filter and an augmented state Kalman filter based on the key parameters, and carrying out the high-robustness state estimation of the vehicle-mounted proton exchange membrane fuel cell. According to the method, the real-time verification of the effectiveness of the Kalman filter based on the standard multi-innovation volume and the dynamic fusion of two estimation results are realized based on the Kalman filter and the fusion estimation based on the effectiveness test and the fuzzy logic, so that an accurate observation mode suitable for the interior of the proton exchange membrane fuel cell system under various conditions is provided; compared with the prior art, the method can effectively improve the estimation performance of the unmeasurable variables under the condition of model mismatch, and is more beneficial to providing support for fuel cell system control.