52 results about "Nonlinear state space model" patented technology

The invention discloses a CAN bus intrusion detection method based on adaptive unscented Kalman filtering, and the method comprises the steps: obtaining real-time message data, and carrying out the preprocessing of the real-time message data, and obtaining a time sequence feature vector; constructing a nonlinear state space model based on the time sequence feature vector; performing unscented Kalman filtering state prediction based on the nonlinear state space model; inputting the time sequence feature vector as an actual observation value, calculating a Kalman gain to correct an unscented Kalman filtering state prediction result, and outputting a state estimation residual error; dynamically updating a process noise covariance matrix through exponentially weighted moving average based on the state estimation residual, and adjusting a measurement noise covariance matrix according to the measurement innovation sequence; calculating the mahalanobis distance of the state estimation residual error, comparing the mahalanobis distance with a self-adaptive anomaly detection threshold value, and judging whether an intrusion behavior occurs or not; and if the abnormal score exceeds a threshold value, triggering a multi-level alarm mechanism, recording a suspicious message and executing a safety protection operation.

The invention discloses a nonlinear aerodynamic damping estimation method based on LSTM, and the method comprises the following steps: 1, building a nonlinear state space model of a structure based on structural response, including a state equation, an observation equation and a relation between nonlinear aerodynamic damping and structural vibration amplitude; 2, performing updating and covariance prediction on response data by using unscented Kalman filtering; 3, correcting the Kalman gain in real time by using a long short-term memory network; 4, performing state updating and covariance updating based on the corrected Kalman gain; 5, training the long-short-term memory network through an unsupervised learning mode, optimizing the filtering performance, and defining a mean square error of a posterior observation predicted value and a real observation value as a loss function; and 6, calculating the nonlinear aerodynamic damping according to the estimated nonlinear aerodynamic damping parameters. The invention further discloses a nonlinear aerodynamic damping estimation system based on the LSTM and a storage medium.

The invention relates to the technical field of digital control switching power supplies, and particularly discloses a switching power supply dynamic response adjusting system based on digital control, which comprises a digital controller, a multi-channel analog-to-digital conversion module, a parameter identification module, a self-adaptive control law adjusting module, a pulse width modulation driving module and a stability monitoring module. On-line identification of key parameters of a system is realized by constructing a nonlinear state space model and combining an extended Kalman filtering algorithm, dynamic compensation is carried out for non-minimum phase characteristics such as a right half plane zero point, and a self-adaptive control law adjustment module optimizes a control strategy in real time according to an updated model. A Lyapunov stability criterion and a predictive control mechanism are introduced, the robustness and response speed of the system are improved, and a stability monitoring module identifies an abnormal state through sliding window variance analysis and peak detection and triggers a fault-tolerant mechanism to guarantee safe operation of the system.

The invention discloses a modeling and optimization control method for a gas supply system of a proton exchange membrane fuel cell, and belongs to the technical field of hydrogen energy power. The method comprises the following steps: establishing a nonlinear state space model of a proton exchange membrane fuel cell gas supply system; determining the optimal oxygen excess ratio of the system under different load currents through experiments, and fitting the optimal oxygen excess ratio into a reference function about the load currents; based on the nonlinear state space model, a model prediction control problem with tracking of the optimal oxygen excess ratio and minimization of the cathode and anode pressure difference as control targets is constructed and expressed as a constrained quadratic programming problem; and decomposing and iteratively solving the quadratic programming problem by adopting an alternating direction multiplier method to obtain the optimal control input of the current control period and act on the system. The method effectively solves the problem that traditional model predictive control is difficult to deploy in real time in a vehicle-mounted controller due to large calculated amount, so that efficient and accurate cooperative control of the proton exchange membrane fuel cell gas supply system is realized.

The invention discloses an ultra-deep well drilling wellbore pressure inversion correction method, and aims to improve pressure prediction precision and well control safety in an ultra-deep well drilling process. The method mainly comprises the following two parts: firstly, establishing a wellbore pressure calculation model, fully considering the change of drilling fluid density along with temperature and pressure based on a mass and momentum conservation equation in a drill rod and an annulus, performing discrete solution on the established control equation by adopting a finite difference method, and performing iterative calculation from a well bottom to a well mouth and from the well mouth to the well bottom, so as to obtain a well pressure calculation model; and acquiring pressure and physical property parameters of each node. Secondly, on the basis of model output, a friction correction factor is introduced, a nonlinear state space model is constructed, an unscented Kalman filter (UKF) algorithm is used for conducting dynamic estimation on the friction correction factor, and online self-adaptive correction on the wellbore pressure model is achieved; according to the method, the accuracy of wellbore pressure calculation is effectively improved, and reliable technical support is provided for ultra-deep well operation.

The invention provides an accurate calculation method and system for the axial pretightening force of an ultrasonic bolt, and the method combines a high-order statistic theory with a nonlinear filtering technology, and solves the problem that the measurement precision of a conventional method is insufficient under the conditions of strong noise and nonlinearity. Particle filtering and Kalman filtering are combined, so that the problem that a traditional Kalman filtering method easily causes sensitivity of an initial value of filtering divergence is solved; a double-Gaussian attenuation model is adopted to fit an echo envelope, the problem that a single Gaussian or single index model cannot adapt to envelope distortion caused by stress change is solved, and a nonlinear state space model is established to adapt to nonlinear echo signal characteristics under dynamic stress; gaussian noise is effectively suppressed by using a third-order cumulant cross-correlation algorithm, the time difference resolution in a low signal-to-noise ratio environment is remarkably improved, and the Gaussian noise and asymmetric interference are effectively suppressed.

The invention discloses a permanent magnet synchronous motor speed regulation control method based on an all-drive system theory. The method comprises the following steps that A, an affine nonlinear standard model is constructed, and whether a PMSM system is an under-actuated system or not is judged based on the affine nonlinear standard model; b, obtaining a pseudo-strict feedback standard system of the PMSM according to the affine nonlinear state space model of the PMSM, and converting the pseudo-strict feedback standard system into an all-drive system model of the PMSM; c, according to the PMSM all-drive system model, constructing an all-drive control law with an expected feature structure and a linear closed-loop steady system; and D, obtaining an expected coefficient matrix through calculation, finally obtaining an all-drive control law with an expected characteristic structure, and carrying out speed regulation control on the permanent magnet synchronous motor by using the all-drive control law. According to the all-drive characteristic of the PMSM nonlinear system, accurate control over the rotating speed ring of the permanent magnet synchronous motor can be achieved, and the dynamic performance and the steady-state performance of the permanent magnet synchronous motor are improved.

The invention discloses a fish growth prediction method based on a depth nonlinear state space model, and belongs to the technical field of intelligent aquaculture. In the method, a fish growth depth state space model is established, a dynamic stage division mechanism is designed, and a different-speed growth gradient and an environmental cumulative effect are fused to construct a stage transition criterion; adaptive switching of model parameters is realized through a gating network, the problem of stage boundary fuzziness caused by traditional fixed threshold division is broken through, biological mechanism equations such as metabolic rate temperature response and density inhibition effect are embedded into a neural network, and interpretable coupling modeling of environment-management factors is realized through regularization constraint of a model parameter space; cross-scale causal association modeling is realized, a metabolism-growth cross-scale causal chain is constructed in combination with biometric prior, and the bottleneck that a traditional black box model cannot associate microscopic physiology and macroscopic growth is broken.

The invention relates to the technical field of automatic control, particularly discloses an intelligent control system for hot galvanizing coating, and aims to solve the problems of large coating thickness fluctuation, high zinc consumption and poor quality stability. The system comprises a multi-modal sensing acquisition module, a dynamic coupling modeling module, a distributed parallel decision module and a multi-target collaborative execution module, by collecting process data in real time, establishing a non-linear state space model, generating a multi-target optimization decision in parallel and outputting a final control instruction based on dynamic weight fusion, accurate control of plating quality and improvement of production energy efficiency are realized.

The invention provides a cooperative control method suitable for a thermoelectric unit, and relates to the technical field of thermoelectric units, and the method comprises the steps: constructing a machine-furnace coupling nonlinear state space model, designing a dynamic heat storage state observer, calculating a heat storage state index in real time to quantify the transient energy profit and loss of the unit, and building a multi-objective optimization function. And a dynamic weighting factor based on the index is introduced, the factor guides the controller to automatically switch strategies under different working conditions according to the priority of the real-time energy state, the intelligent balancing load response speed and the main steam pressure stability, and finally, a nonlinear model predictive control algorithm is adopted to solve in a rolling time domain, so that the real-time energy state is obtained. According to the method, the optimal steam turbine control valve and fuel quantity control increment is obtained, decision-making level deep cooperation of a turbine-boiler system is effectively achieved, the problem of divergence control under deep peak regulation is solved, the frequency modulation potential of a unit is released to the maximum extent on the premise that absolute safety of pressure is guaranteed, and the response rate of a power grid is increased.

The invention belongs to the field of artificial intelligence, particularly relates to a meteorological industry sensor intelligent calibration and calibration and full-process automatic detection method, and aims to solve the problems that traditional calibration statics are lagged, abnormity depends on manual interpretation, operation and maintenance response is slow and the like. According to the method, a multi-source heterogeneous sensor network is constructed, data are synchronously acquired, a personalized error compensation function is established for each sensor based on a nonlinear state space model, and online adaptive calibration is realized through edge nodes; and carrying out multi-scale anomaly detection in combination with a depth time sequence convolutional neural network and an isolated forest algorithm. According to the scheme, concurrent processing of ten thousand people-level stations is supported, the single-point delay is lower than 8 seconds, the calibration deviation is smaller than 0.5%, the anomaly recognition accuracy is higher than 97%, equipment failure can be predicted 30 days in advance, the operation and maintenance cost is reduced by 40%, the numerical forecasting initial field precision is improved by 12%, and a high-reliability, full-automatic and intelligent meteorological sensing data quality management and control system is comprehensively achieved.

The invention provides a state space model prediction optimization control system and method in a water-light complementary system, and relates to the field of power system automatic power generation control and renewable energy source grid connection. For uncertainty caused by photovoltaic output and load fluctuation in the water-light complementary system, a prediction controller based on a state space model is constructed. The control method comprises the following steps: establishing a nonlinear state space model of the water-light complementary system, and performing linearization processing at a steady-state point to obtain a discrete state space prediction model; on the basis of future dynamics of the model prediction system, designing a cost function including prediction error compensation, and adopting a quadratic programming method to solve an optimal control sequence in an online rolling manner; finally, the first component of the control sequence acts on the hydroelectric generating set speed regulation system, and frequency stability and power balance are achieved. Through model prediction, rolling optimization and feedback correction mechanisms, the adaptive capacity of the system to photovoltaic fluctuation and load disturbance is remarkably improved, and the robustness and dynamic response performance of the system are enhanced.

Proposed is a new posterior distribution estimation technique. This estimating device comprises at least one memory and at least one processor, wherein, when estimating a posterior distribution of a state at a second time point based on observed values up to the second time point, in a case of a non-linear state space model in which the state distribution at the second time point, evolved over time on the basis of the state at a first time point, is described by a linear Gaussian distribution, and the distribution of observed values at the second time point, based on the state at the second time point, is described by a non-linear Gaussian distribution, the at least one processor uses a model that estimates a linear Gaussian distribution and that has parameters that are optimized to maximize a lower bound on a log-likelihood, to calculate a posterior distribution of the state at the second time point based on the observed values up to the second time point.

The invention discloses a multi-underwater robot cooperative formation control method, and belongs to the technical field of underwater robots. According to the method, the state quantity of the optimal nonlinear state space model of each AUV is determined by adopting a Hungary method, so that the problems of track intersection and collision between different AUVs in the formation transformation process can be avoided. Besides, under the condition that thrust is limited, a formation control objective function including a relative pose error term and a safety constraint penalty term is adopted, the distance between different AUVs is dynamically adjusted, and active avoidance and safe formation transformation are achieved.

This application discloses a method and apparatus for variable load control of solid oxide fuel cells (SOFCs), relating to the field of power load control. The method includes establishing a nonlinear state-space model of the SOFC based on its heat flow model; transforming the nonlinear state-space model into a standard state-space model using the Jacobian matrix linearization method; and performing variable load control using a model predictive control algorithm based on the power output objective function, constraints, and the standard state-space model. This invention establishes a cross-scale heat flow model based on the overall heat transfer process of the SOFC, selects appropriate state variables and inputs, and establishes a nonlinear state-space model of the SOFC guided by variable load. The standard state-space model is obtained using Jacobian matrix linearization, and the power is adjusted under multiple constraints using a model predictive control algorithm to meet the expected load changes.

The application discloses a kind of drilling tool gravity tool face angle dynamic measurement method and device. With gravity tool face angle and gyroscope drift as state variable, with accelerometer y-axis and z-axis measurement value as observation, fusion drilling tool measurement matrix and system nonlinearity establishes nonlinear state space model, linearizes nonlinear state space model at current working point;With central symmetric polytope to the bounded component of state variable initial value, process disturbance and observation noise, with random vector to the random component of system state variable initial value, process disturbance and observation noise, establish hybrid Kalman filter observer;Through forgetting factor attenuation hybrid Kalman filter prior covariance matrix historical information, according to hybrid Kalman filter observer hybrid prior covariance matrix, observation matrix and hybrid new information covariance matrix determine hybrid Kalman filter optimal gain, utilizes the hybrid Kalman filter observer established to obtain gravity tool face angle.

The invention provides a heavy-duty commercial vehicle low-adhesion road surface automatic emergency braking system control method, a commercial vehicle and a medium, and belongs to the technical field of commercial vehicle intelligent driving. According to the method, data of a wheel speed sensor, a GPS, an IMU and a camera are fused through dynamic weight distribution; establishing a vehicle nonlinear state space model, and estimating a vehicle state and a road adhesion coefficient in real time by adopting extended Kalman filtering; calculating the collision time TTC according to the estimated state, and dynamically adjusting the alarm and brake threshold according to the road adhesion coefficient; and finally, early warning or braking is triggered by comparing TTC with an adjusted threshold decision. Through multi-sensor fusion and adaptive weight adjustment, the estimation precision of the vehicle state and road surface recognition on the low-adhesion road surface is improved. Based on a dynamic threshold adjustment mechanism of a road adhesion coefficient, the AEB system can adapt to different adhesion conditions, and the active safety and the braking reliability of the commercial vehicle when the commercial vehicle runs on a low-adhesion road surface are improved.

The invention discloses a force-displacement synchronous introspection sensing method for the tail end of an electrothermal micro-actuator, and the method comprises the steps: defining a physical input variable and an output variable of a state space model of the electrothermal micro-actuator, and selecting an intermediate variable which can be directly or indirectly measured as a state variable; constructing a nonlinear state equation and a nonlinear output equation, and selectively adding nonlinear terms; reasonable driving voltage excitation signals and tail end load force excitation signals are selected, and data of input variables, output variables and state variables responded in the whole process are synchronously collected; performing parameter identification on the nonlinear state space model by adopting a nonlinear dynamic sparse identification algorithm to obtain a continuous time state equation and an output equation of the system; and converting the continuous time state equation and the output equation into a discrete time state equation and a discrete time output equation, and synchronously estimating the loading force and the output displacement of the tail end of the micro-actuator on line by using state variable data and driving voltage data which are acquired in real time.

The invention discloses a signal denoising and carrier synchronization method and device of a continuous wave mud pulse system, and belongs to the field of measurement while drilling in petroleum drilling. The method comprises the following steps: S1, preprocessing a collected mud pulse signal; s2, adjusting a noise covariance matrix of an adaptive Kalman filter in real time based on an expectation maximization algorithm, and reconstructing and filtering harmonic waves of each order of pump noise through a pump noise linear time-invariant space state model by using the adjusted adaptive Kalman filter; s3, filtering random noise by adopting a wavelet threshold denoising method; s4, performing frame synchronization on the denoised signal by using the m sequence; s5, demodulating the mud pulse signal by adopting a BPSK (Binary Phase Shift Keying) orthogonal demodulation technology; and S6, adjusting a noise covariance matrix of the unscented Kalman filter in real time based on an expectation maximization algorithm, and estimating and compensating the phase offset by using the adjusted unscented Kalman filter through the nonlinear state space model of the phase offset and the frequency offset.

The invention discloses a motor system simulation method under the condition of data anomaly, solves the problem of inaccurate parameter estimation of a motor system model under the conditions of data anomaly and non-Gaussian interference noise of a nonlinear state space model, and belongs to the technical field of motor system modeling. The method comprises the following steps: establishing a nonlinear state space model of a motor system; under a reasoning framework, each component of measurement noise in the nonlinear state space model obeys independent univariate skew distribution, modeling of non-Gaussian noise is realized, hierarchical modeling of noise is realized by using auxiliary variables, and probability distribution of unknown parameters in the nonlinear state space model is determined; identifying unknown parameters in the nonlinear state space model based on a variational Bayesian reasoning framework and the probability distribution of the unknown parameters to obtain posterior distribution of the unknown parameters; and simulating the motor system by using the non-linear state space model of which the unknown parameters are identified.

The invention belongs to the technical field of wind power prediction, and particularly relates to a method for predicting ultra-short-term wind power based on an improved UKF algorithm. According to the method, a square root UKF, normalized innovation square gating and self-adaption, multi-information joint updating and multi-model interaction technologies are fused on the basis of the UKF. The method comprises the following steps: establishing a nonlinear state space model, generating a propagation sigma point by SR-UKF to complete time updating, acquiring, measuring and executing NIS gating, fusing and innovating a multi-information strategy and self-adapting to a noise covariance, constraining a projection to ensure that a physical boundary and the covariance are positive definite, fusing multiple working condition probabilities by IMM, and outputting a prediction result and uncertainty information. According to the method, linearization errors are avoided, and prediction robustness is improved; the problems of phase lag, unstable numerical value and the like are solved through multiple technologies respectively, second-to-minute level accurate prediction is achieved, and meanwhile deliverable capacity information capable of being directly used for control and performance is output.

The present application relates to the technical field of battery monitoring, and particularly relates to a power battery side reaction real-time monitoring method and system based on an electrochemical model. The method comprises the following steps: S1, constructing an enhanced electrochemical model, the enhanced electrochemical model comprising a coupled kinetics equation describing a lithium precipitation side reaction and a SEI film growth side reaction inside a battery; S2, performing numerical discretization processing on the enhanced electrochemical model, and converting the enhanced electrochemical model into a nonlinear state space model, wherein a state vector at least comprises state variables representing a negative electrode solid-phase lithium concentration, a positive electrode solid-phase lithium concentration, an electrolyte lithium ion concentration and a SEI film thickness; S3, taking current and voltage collected by a battery management system in real time as input, performing a prediction-correction cycle through a state observer, and outputting an optimal estimation value of the state vector in real time; S4, extracting a quantitative index for representing a battery safety state from the optimal estimation value in real time; and S5, executing a hierarchical safety early warning strategy according to the quantitative index.

The application discloses a nonlinear aerodynamic damping estimation method based on an LSTM, and comprises the following steps: step one: constructing a nonlinear state space model of a structure based on a structural response, including a state equation, an observation equation and a relationship between nonlinear aerodynamic damping and a structural vibration amplitude; step two: updating and covariance prediction of response data by using an unscented Kalman filter; step three: real-time correction of Kalman gain by using a long short-term memory network; step four: state updating and covariance updating based on the corrected Kalman gain; step five: training the long short-term memory network in an unsupervised learning mode, optimizing filter performance, and defining a mean square error of a posterior observation prediction value and a true observation value as a loss function; and step six: calculating nonlinear aerodynamic damping according to the estimated nonlinear aerodynamic damping parameters. The application further discloses an LSTM-based nonlinear aerodynamic damping estimation system and a storage medium.

The application aims to provide a real-time online diagnosis method for engine valve clearance abnormal failure based on DREKF, belonging to the field of engine diagnosis. The method comprises the following steps: real-time acquisition of engine cylinder head vibration signal, establishment and simplification of cylinder head vibration signal mathematical model and determination of undetermined parameters; selection of appropriate state variables to construct cylinder head vibration signal state space equation and verification of its observability; design of double-rate extended Kalman filter optimal estimation method, input of vibration signal into DREKF, and obtaining of cylinder head vibration signal estimated value; simultaneous calculation of process derivative based on DREKF estimated vibration signal and derivative threshold value, and realization of online real-time fault diagnosis of the engine valve clearance according to the valve clearance state judgment criterion. The application establishes a new nonlinear state space model of engine cylinder head vibration signal, is suitable for engine cylinder head vibration signal observation under different speeds and different loads, and realizes online real-time diagnosis of the valve clearance state.

The application relates to a nonlinear anti-saturation robust control method of a flexible satellite attitude system based on a dynamic observer, and comprises the following steps: S1, establishing a flexible satellite attitude system, and converting the flexible satellite attitude system into a corresponding nonlinear state space model; S2, constructing a nonlinear dimension reduction dynamic observer according to the nonlinear state space model, designing a nonlinear state feedback robust controller based on the nonlinear dimension reduction dynamic observer, constructing a nonlinear anti-saturation compensator based on the nonlinear state feedback robust controller, and obtaining a nonlinear closed-loop control system which can simultaneously process state unmeasurability, convex polyhedral uncertainty, external disturbance and actuator saturation problems; S3, respectively deriving SOS solvability conditions under the conditions of ignoring and considering actuator saturation; and S4, respectively solving the nonlinear dimension reduction dynamic observer and the state feedback robust controller, and the nonlinear anti-saturation compensator under the conditions of ignoring and considering actuator saturation.

This invention discloses a nonlinear predictive function control method for a fractional-order turbine regulating system, belonging to the field of automatic control technology. First, a sudden load model of the turbine regulating system is introduced, incorporating fractional calculus and dynamic turbine transfer coefficients. Then, based on the series characteristics of the turbine regulating system, a predictive function controller based on a time-varying nonlinear state-space model is designed, effectively improving the unit's speed and stability. The fractional-order model part employs an improved Oustaloup approximation method and an optimal model order reduction algorithm, enabling efficient computer simulation calculations.

The invention provides a lithium ion battery state estimation method based on an advanced mobile vision field. The method comprises the following steps: establishing an equivalent circuit model: establishing a lithium ion battery equivalent circuit model containing a hysteresis effect and temperature related parameters; constructing a nonlinear state space model based on the equivalent circuit model, wherein the state space model comprises a state equation for describing dynamic behaviors of the system and an observation equation for reflecting voltage response; an event trigger linearization mechanism (ETR) is introduced into a mobile vision field estimation (MHE) optimization framework; dynamically adjusting the weight of the cost function by combining a Gaussian-Newton iteration (GN) optimization method; and a high-precision SOC estimation result is output by using real-time measurement data through complete modeling, prediction and optimization processes. On the basis that the advantages of an MHE optimization structure are reserved, battery physical nonlinear factors are further fused, and the SOC estimation method with temperature adaptability solves the problem of estimation deviation caused by the fact that a traditional MHE simplified model ignores physical characteristics.

The application relates to the technical field of automatic control, and particularly discloses a hot galvanizing coating intelligent control system, which aims to solve the problems of large coating thickness fluctuation, high zinc consumption and poor quality stability. The system comprises a multi-modal sensing acquisition module, a dynamic coupling modeling module, a distributed parallel decision module and a multi-target collaborative execution module. Through real-time acquisition of process data, a nonlinear state space model is established, multi-target optimization decisions are generated in parallel, and the final control instruction is output based on dynamic weight fusion, so that the accurate control of coating quality and the improvement of production energy efficiency are realized.

The invention provides a train operation state estimation method and system fused with a remainder extended Kalman filter, and belongs to the technical field of rail transit train operation management based on deep learning. Predicting based on the nonlinear state space model of the train to obtain an intermediate state estimation result of the train; a deep neural network module based on an LSTM structure takes the intermediate state estimation result, the error covariance, the Kalman gain and the system residual error as input features, and outputs a group of optimized state estimation results through nonlinear combined learning model error distribution; and carrying out weighted fusion on the intermediate state estimation result and the optimized state estimation result to obtain a final state estimation result. According to the method, the precision of system state estimation is improved, the nonlinear modeling capability of the system is enhanced, the complex nonlinear relationship in the system can be adaptively captured and learned, and the defects of a traditional filter in processing the nonlinear system are overcome.

The invention discloses a dense-medium suspension density robust soft measurement method based on mixed distribution, which comprises the following steps: inputting valve opening data into a dense-medium suspension density robust soft measurement model obtained by training, and outputting target dense-medium suspension density data; training to obtain a dense-medium suspension density robust soft measurement model: constructing a nonlinear state space model based on density control loop characteristics of a dense-medium coal separation process density loop system; historical operation data of the densimeter is collected, and an identification data set is constructed; the nonlinear state space model comprises a state transition equation and an output measurement equation; modeling the noise characteristics by using a mixed probability distribution theory to obtain a nonlinear state space robust probability model; an expectation maximization algorithm is adopted to carry out joint parameter identification and state estimation, an optimal parameter combination of a nonlinear state space robust probability model is determined through iterative optimization, a dense medium suspension density robust soft measurement model is obtained, and the problem that a traditional soft measurement method is poor in robustness is effectively solved.