一种残差模糊匹配的非线性目标跟踪改进算法

为解决目标跟踪中因系统滤波初值不准确和噪声统计特性未知引起标准非线性卡尔曼算法估计误差变大问题,该文提出一种基于残差的模糊自适应(RTSFA)非线性目标跟踪算法。在确定采样型滤波基本框架的基础上,给出了在线性化误差约束条件下高斯权值的积分一般形式,并利用李雅普诺夫第二方法证明了该算法估计误差有界收敛的充分条件。进一步构建自适应噪声协方差矩阵在线估计噪声特性,并引入Takagi-Sugeno模型和量测椭球界限规则选择噪声估计器调节因子,有效提高了算法的收敛速度和滤波精度。通过滤波初值信息不明和量测噪声时变的纯方位目标跟踪模型,验证了非线性目标跟踪算法具有更好的跟踪精度和更强的鲁棒性。

An Improved Algorithm for Nonlinear Target Tracking Based on Residual Fuzzy Matching

A nonlinear target tracking algorithm based on residual Takagi Sugeno fuzzy adaptation (RTSFA) is proposed in order to solve the problem that the estimation error of the standard nonlinear Kalman algorithm increases due to the inaccurate initial value of system filter and the unknown statistical characteristics of the noise in the target tracking. Based on the basic framework of deterministic sampled filtering, the general form of Gaussian weight integral under linearization error constraints is given, and the sufficient condition for the bounded convergence of the estimation error of the algorithm is proved by the Lyapunov''s second method. Furthermore, the on line estimation noise characteristics of the adaptive noise covariance matrix are further constructed, and the Takagi Sugeno model and the measurement ellipsoid boundary rule are introduced to select the noise estimator adjustment factor, which can effectively improve the convergence speed and the filtering accuracy of the algorithm. It is verified that the nonlinear target tracking algorithm has better tracking accuracy and stronger robustness by using the pure azimuth target tracking model with unknown initial value information and time variation measurement noise.