Fuzzy Gaussian sum particle filtering method and device as well as target tracking method and device

Abstract

The invention discloses a fuzzy Gaussian sum particle filtering method and device as well as a target tracking method and device. The fuzzy particle filtering method comprises the following steps: a state posterior probability density function, an observation noise probability density function and a state noise probability density function of a last target moment are established with a Gaussian sum method, the prediction probability density function of a target state of the current moment is acquired with Gauss-Hermite quadrature rules and the Monte Carlo principle, the particle weight and integral point weight of the current target observation moment are acquired with the fuzzy clustering principle, the weight of each Gaussian term is calculated for calculation of the mean and covariance of each Gaussian distribution, resampling is performed on the Gaussian terms, G Gaussian terms with larger weights are acquired, then the state posterior probability density function of the current target observation moment is acquired with the Gaussian sum principle, and particle filtering is finished. With adoption of the scheme, the filtering accuracy and estimation performance of a target state can be improved effectively.

Description

Technical field [0001] The present invention relates to the field of nonlinear filtering, in particular to a fuzzy Gaussian and particle filtering method and device, and a target tracking method and device. Background technique [0002] During the movement of targets such as airplanes, aviation vehicles, and vehicles, it is often necessary to estimate the real-time state of the target in order to track the target. The motion system model of the target such as airplane is generally a nonlinear stochastic system. Non-linear filtering technology is a common method for state estimation in non-linear stochastic systems. [0003] According to different application backgrounds, the prior art nonlinear filtering technologies are mainly divided into two categories: The first category is for state estimation problems in nonlinear Gaussian environments, such as extended Kalman filter (EKF) and unscented Kalman filter (UKF), Integral Kalman Filter (QKF), Truncated Unscented Kalman Filter (IUK...

AI Technical Summary

Problems solved by technology

The second category is for state estimation problems in nonlinear non-Gaussian environments, such as Gaussian sum filter (GSF), Gaussian and integral Kalman filter (GS-QKF). This type of Gaussian sum method mainly uses multiple mixed Gaussian The posterior probability density function of the state is approximated as a single Gaussian function. However, similar to the above-mentioned EKF and other methods, such Gaussian sum methods must be linearized. For strongly nonlinear non-Gaussian systems, the filtering accuracy of such filters is not as good as is not high, and the number of Gaussian mixture terms of the filter grows rapidly with time

[0004] In addition, in order to deal with nonlinear non-Gaussian noise, the existing technology also adopts another nonlinear filtering method: Gaussian and particle filtering methods, which are suitable for large-scale passive sensor systems with limited observation information, high data loss rate, and non-periodic , non-linear, non-Gaussian observation data, a class of Gaussian and particle filter methods used in the prior art simply use the state transition function for particle sampling when time is updated, and the diversity and accuracy of the particles are poor. Particles cannot effectively represent the posterior probability distribution of the target, thereby reducing the performance of the particle filter

Images