51 results about "Linearization error" patented technology

The invention discloses a target predicting and tracking method based on a probability graph model, which builds a target tracking model by a probability graph model modeling method, estimates the target position by the particle filtering prediction algorithm, and infers the possible positions of the target at any time by a distributed particle filtering tracking method, thus realizing the target tracking. The invention is not restricted by linearization error and Gaussian noise assumption, does not need to assume the distribution form of sample points, can deal with any density distribution and flexibly adapt to non-linear dynamic models and multimodal observation models of target tracking. The method can obtain higher overall predicting degree of accuracy and show better target tracking performance.

The invention discloses a Stirling interpolation polynomial approximation based ellipsoidal set-membership filtering method. The method includes: firstly utilizing Stirling interpolation polynomial to obtain linear approximation of a nonlinear state equation, constructing an uncertain interval of system state variable, using a second order difference operator to obtain a linear error bound, constructing a virtual noise outsourced ellipsoid, using the current state variable estimate value to predict a system state variable prediction error boundary of the next moment, carrying out system state variable update operation according to the observation vector, then utilizing linear ellipsoid set-membership filtering steps to carry out estimation calculation of system state variable and calculation of system state variable's estimation error ellipsoid, thus completing the estimation calculation task of system state variable. The method provided by the invention utilizes Stirling interpolation polynomial approximation calculation to realize ellipsoid set-membership filtering, effectively reduces the calculation amount, enhances the computational efficiency, and improves the calculation precision of extended set-membership filtering algorithm.

The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to an unmanned reconnaissance aerial vehicle synchronous positioning and picture compositing method based on a self-adaption EKF. By means of the method, a graph model, a coordinative system, a feature model, an unmanned reconnaissance aerial vehicle movement model and a sensor measurement model of an unmanned reconnaissance aerial vehicle synchronous positioning and picture compositing system are established. Synchronous positioning of the unmanned reconnaissance aerial vehicle and design of a map establishment method frame are achieved through the EKF. The problems that due to the establishment inaccuracy of the unmanned reconnaissance aerial vehicle movement model and an observation model and the linearization of non-linear models, the models are not matched are researched. The uncertainty caused by establishment inaccuracy and linearization errors of the models is modeled into process noise and observation noise of the system with a time-varying mean and a covariance matrix, and the synchronous positioning and picture compositing method based on the self-adaption EKF with a time-varying noise estimator is designed. Statistical features of the process noise and observation of the system are estimated in real time and modified, and therefore model uncertainty is reduced and the estimation progress is improved.

The invention provides a mixed integer nonlinear programming model for solving the combination problem of a hydroelectric generating set, and belongs to the field of hydroelectric scheduling operation. The mixed integer nonlinear programming model for solving the combination problem of a hydroelectric generating set is characterized by utilizing a polynomial fitting technology to process the complicated power station nonlinear relation, providing a quartic equation polynomial fitting method of water level-capacity of reservoir-tailwater level-out flow, and based on a quadratic polynomial of head loss of the hydroelectric generating set, describing the generating head of the hydroelectric generating set as a multivariate quartic function of capacity of reservoir and flow; considering aboutthe influence of the head on the output, providing a binary quadratic polynomial method of accurately representing the output head and the generating flow of the hydroelectric generating set, introducing 0-1 variable, solving the startup-shutdown problem of the hydroelectric generating set, obtaining the head-flow-output relation three-dimensional fitting surface of the hydroelectric generating set, and satisfying the fine calculation demand of the output of the hydroelectric generating set under the variation head; and realizing calculation with high efficiency by means of a mature MINLP solver. The mixed integer nonlinear programming model for solving the combination problem of the hydroelectric generating set fully considers the complicated nonlinear relation of the hydroelectric systemand has the advantages of reducing the linear error and significantly improving the solution accuracy, compared with a classic MILP model. Besides, the obtained set combination and load distributionresult can preferably reflect the practical operation situation of the hydropower station.

The invention provides a space rendezvous system gain scheduling control method with linearization errors taken into consideration, and relates to a spacecraft orbit rendezvous gain scheduling control method. The space rendezvous system gain scheduling control method solves the problem that due to the fact that according to an existing aircraft orbit rendezvous control method, input saturation and parameter uncertainty caused by the linearization errors are ignored, time consumption is high when an aircraft orbit rendezvous task is completed. According to the space rendezvous system gain scheduling control method, the parameter uncertainty caused by the linearization errors is taken into consideration and endowed with exact meanings, an aircraft orbit rendezvous relative movement model is established, a gain scheduling controller of an aircraft orbit rendezvous is then designed, the aircraft orbit rendezvous is controlled through the gain scheduling controller, and the rendezvous task is completed. The space rendezvous system gain scheduling control method is mainly used for controlling the aircraft orbit rendezvous.

The invention discloses a multi-target hierarchical prediction control method based on a continuous stirred tank reactor. According to the control method, a double-layer prediction control algorithm is introduced and changes the situation that in the control process of the CSTR, only a dynamic control target is controlled, and economic indexes are ignored. In the control process, a two-level hierarchical prediction control structure is adopted, an expansion Kalman observer is introduced in a dynamic control layer to perform unbiased estimation on the state, a multi-target hierarchical optimization is adopted in a steady-state optimization layer, and important optimization targets are met preferentially. Meanwhile, a finite element orthogonal configuration method is adopted in a whole control system so that a nonlinear optimization control proposition can be converted into a nonlinear programming problem, a model equation serves as a constraint condition, values on configuration points serve as prediction values, a prediction model can be obtained without linearization or step response, and therefore linear errors can be avoided, and integral objects need to be independently processed.

The invention discloses a battery SOC estimation method based on HCKF. On the basis of a battery electrochemical model, parameters are identified through a least square method; the CKF is used as a determined sampling type filtering algorithm; when the nonlinear equation is processed, a point set is generated according to the mean value and the covariance of the prior probability density distribution of the system state and a certain sampling strategy, then each sampling point in the point set is directly subjected to nonlinear propagation, and finally the mean value and the covariance of theposterior probability density distribution of the system state are calculated through weighted summation. According to the method, the nonlinear equation does not need to be linearized, linearizationerrors are eliminated, a Jacobian matrix in the EKF does not need to be calculated in the iterative process of the filtering algorithm, and the method is easier to use in practice; an HCKF algorithm combining a CKF and an H _ infinity filter is proposed to be used for estimating the SOC; the situation that SOC estimation is not accurate enough when battery model errors, unknown measurement noise characteristics and other problems exist is effectively avoided; and the robustness is greatly improved.

The invention discloses a flux linkage full-rank identification method for a permanent magnet of a PMSM. According to the method, an unscented-Kalman-filter-algorithm-based parameter sensitivity analysis for the flux linkage identification precision of a permanent magnet of a PMSM is carried out; a permanent magnet flux linkage full-rank identification method of different rotating speed zones of a PMSM drive system is determined. With the unscented-Kalman-filter-algorithm-based permanent-magnet flux linkage full-rank identification method of different rotating speed zones of a PMSM drive system, influences on the permanent magnet flux linkage identification precision by the system measurement noises, PMSM parameter changing, and a degenerate-rank phenomenon of an identification model can be eliminated; linear error occurrence and complicated Jacobian matrix calculation during the state identification process of the traditional extended Kalman filter algorithm can be avoided; and the algorithm implementation difficulty can be reduce. Meanwhile, the number of to-be-identified parameters can be reduced at a high-speed operating zone in a PMSM drive system; the operand of the identification algorithm can be reduced; and reasonable consideration of the identification precision and the identification speed can be realized.

The invention discloses a control system and a control method for mining the acceleration potential of an aero-engine. The control system comprises the aero-engine, an optimization controller, a full authority digital electronic controller and an engine non-linearity real-time mathematical model, wherein a sensor used for testing the rotating speed, temperature and pressure of the engine is installed on the aero-engine, and the full authority digital electronic controller controls the aero-engine through an electromagnetic valve component. On the basis of avoiding model error brought by linearization treatment of an engine model, compared with the existing control algorithm at present, the control method disclosed by the invention has the advantages that after a designed optimizing control algorithm introduces security constraints on the basis of ensuring the optimizing control effect, so that the working security of the engine in the acceleration process can be greatly improved; compared with the traditional control method, the control method disclosed by the invention can be used for avoiding the linearization error of the engine model, and greatly improving the confidence coefficient of the system optimization effect.

The invention discloses a multi-point linearized probability energy flow method of an integrated energy system with electricity converting to natural gas. The method includes the following steps: firstly proposing an energy hub model containing wind electricity, thermoelectricity combined supply, and electricity converting to natural gas; then conducting quantitative analysis on precision of linearized energy flow equations of a power grid and a natural gas net, and subsequently making a conclusion that the precision of the linearized energy flow of the natural gas net is explicitly lower than that of the power grid. Based on the conclusion, the invention provides a method for calculating the multi-point linearized energy flow. At the same time, the method takes into consideration of the relativity between electricity, natural gas and thermal load and wind velocity, and based on a probability density distribution function of undetermined energy flow injection, adopts the monte carlo simulation method in generating random samples. For each random sample, the method adopts the multi-point linearization method in calculating the energy flow. According to the invention, the method has high calculation efficiency, and at the same time has linearization errors which are within an acceptance scope.

The invention discloses an electric control air suspension vehicle body height adjusting method based on model predictive control. The method comprises the following steps: establishing and linearizing a nonlinear vehicle height adjusting model of a quarter of an electric control air suspension to serve as a prediction model; observing the model linearization error and the road surface interference by using an extended state observer and feeding back to the prediction model; according to the target vehicle body height, the actual vehicle body height and the prediction model, calculating the required air mass flow through model prediction control and then converting into opening and closing signals of an electromagnetic valve through PWM control; designing an adaptive event triggering mechanism, and judging whether optimization calculation of model prediction control is carried out or not at each sampling moment according to an adaptive threshold and a triggering condition. Aiming at the characteristics of strong non-linearity, susceptibility to road surface interference and large calculation amount of a control algorithm, the invention provides the electric control air suspension vehicle body height adjusting method based on model predictive control, so that the speed, precision and robustness of vehicle body height adjustment are improved.

The invention discloses a mine personnel safety state estimation method based on set membership filtering under incomplete measurement. The method fully considers the incomplete measurement of a wireless sensor and unknown but bounded noise in a mine environment and constructs a set membership filter based on the incomplete measurement. The objective of the invention is to effectively process nonlinear errors in a mine operating personnel positioning system. According to the method, a system measurement equation is subjected to linearization processing, a scaling matrix and an unknown matrix are adopted to represent a linearization error, and a particle swarm algorithm is utilized to process the scaling matrix such that the linearization problem of the system is effectively solved, and theaccurate positioning of mine operating personnel is realized. Moreover, various filtering parameters are determined through positive semidefinite programming, and a recursive filtering algorithm is proposed to calculate a minimum state estimation ellipsoid containing the real state of the mine operating personnel such that the reliability of the mine operating personnel positioning system is improved.

The invention discloses a self-adaptive linearization probabilistic power flow calculation method containing high-proportion wind power grid connection, and the method is used for solving the problemin power system power flow calculation considering high-proportion wind power and load strong random disturbance. The method comprises the steps that firstly, a Gaussian mixture model is adopted for uniformly depicting source-strong load randomness, and a source-load uncertainty model is established; next, an input variable correlation sample is generated by random simulation; then, self-adaptivelinearization processing is carried out on the power flow equation, and a random fluctuation range adaptive multi-area division is realized by adopting an iterative algorithm, and linearization is carried out in the division region so as to reduce the global linearization error of the power flow, and an analysis method of a self-adaptive linearization semi invariable method is adopted to obtain semi invariants of each order of the state variables after regional integrative recombination; and finally, state variable probabilistic distribution fitting is carried out through C type Gram-Charlierseries. By adoption of the method, the source-strong load randomness and correlation can be processed effectively, more accurate power flow distribution can be acquired, and advantages of accurate result and convenient realization are realized, so that the method is applicable to analysis and evaluation of the probabilistic risk in the high-proportion wind power grid connection.

The invention provides a method for estimating an external ballistic flight state of a cannonball. The method comprises the steps of modeling a system, improving a filtering algorithm and selecting the number of sample points. The improved filtering algorithm is adopted for estimation of external ballistic parameters of the cannonball; a point-mass trajectory equation serves as a state model of the system; and a radar measurement equation serves as a measurement model of the system. The estimation method adopts an improved nonlinear filtering algorithm. According to the method for estimating the external ballistic flight state of the cannonball, a state estimation error can be kept near a certain steady-state range after filtering is started for 5s, so that the convergence effect is good.The estimation method has the advantages that a conventional estimation method is remarkably improved, the measurement precision of external ballistic parameters of a missile rocket is improved, the introduction of a linearization error is avoided under the condition of ensuring the filtering precision and the algorithm calculation amount, and the filtering precision is improved, so that an important basis is provided for performing accurate ballistic control on the cannonball.

The invention discloses a discrete sliding mode trajectory and speed tracking control method for a high-speed train. The method comprises the following steps: constructing a nonlinear train model of the train, and assigning estimation parameters; with displacement and control input quantities as the input and output data for model optimization, and discretizing and linearizing the nonlinear trainmodel so as to obtain a discrete linear train model with optimized model parameters; designing an extended Kalman filter oriented to the discrete and linearized model to obtain filtered train displacement and an estimated speed; designing a high-speed train trajectory tracking controller by adopting a discrete sliding mode method, and designing an interference observer to estimate model uncertainty and linearized errors so as to improve train tracking precision. According to the invention, train model parameters are estimated by adopting a chaos optimization method, the nonlinear train model is discretized and linearized, and then train displacement and speed estimation is designed by adopting a discrete extended Kalman filtering method, so measurement noise interference of train displacement is relieved, and the problem of speed sensor faults is solved.

The invention discloses a target tracking method based on progressive unscented Kalman. The method considers the truncation error problem in a measurement linearization process and puts forwards a progressive unscented Kalman filter algorithm to better handle linearization error and numerical calculation error increasing problems existing in the measurement linearization process. Firstly, a homotopy function for a system state is constructed according to a Bayes rule; through carrying out iteration on a measurement updating process, the current observation information is introduced progressively, and further, the posterior state of the system is obtained; and besides, an innovation judgment condition is introduced during the progressive process, the estimated errors of the system during the progressive process are ensured to be bounded, and the stability of the filter is ensured. Compared with the existing target tracking method, the method thoroughly considers linearization error influences, and on the premise of ensuring the calculation complexity, the target tracking precision and the robustness are improved.

The invention discloses a target tracking method based on distributed matrix weighted fusion Gaussian filtering, and the method comprises the steps: firstly carrying out the local state estimation through Gaussian filtering, then approaching an error cross covariance between local estimation through employing a statistical linear regression method, and finally obtaining an optimal parameter under a maximum likelihood criterion through solving an optimization problem, and adjusting the error cross covariance matrix. According to the method, the performance loss caused by linearization errors in statistical linear regression is reduced, and the target tracking precision is improved.

The invention discloses a pseudo-satellite indoor positioning method based on a UKF algorithm. The method comprises the steps that (1) a pseudo-satellite double-difference pseudo-range observation model is constructed; and (2) the UKF algorithm is used to carry out indoor positioning on a pseudo-satellite. According to the invention, the pseudo-satellite double-difference pseudo-range observationmodel is constructed to eliminate the influence of receiver clock difference, satellite clock error, thermosphere delay and troposphere delay in pseudo-satellite pseudo-range measurement values; for the problem of a large linearization error generated by an independent networking pseudo-satellite system which uses an EKF algorithm, the linearization error of the observation model is analyzed; theinfluencing factor of the linearization error is the error of the distance between the pseudo-satellite and a receiver and the position of the receiver; when the coordinate error of the receiver is larger and the distance between the satellite and the receiver is smaller, the linearization error is larger; in pseudo-satellite indoor positioning, the UKF algorithm is used for positioning to avoid the linearization error; and when the coordinate error of the receiver is relatively large, faster convergence speed and higher indoor positioning accuracy than the EKF algorithm can be acquired.

The invention discloses a power flow model linearization error minimization method based on variable space optimal selection. The method mainly comprises the following steps: 1) establishing a linearpower flow model based on a power system node space by utilizing an finite-order Taylor expansion method, (2) establishing an optimal variable space selection model on the basis of the linear power flow model, and (3) inputting real-time power system parameters into the optimal variable space selection model to obtain an optimal variable space. The optimal variable space of the method proposed bythe present invention can effectively reduce the linear error when the operating state of the system fluctuates in a large range.