基于统计线性回归的粒子滤波方法

针对非线性、非高斯系统的状态估计问题,该文提出了一种基于统计线性回归的粒子滤波算法.在该算法中,首先对非线性函数基于统计线性回归展开,并利用高斯积分估计回归系数,依此产生重要性密度函数.该密度函数融入了最新的观测信息,扩大了与系统真实后验密度的重叠区域.理论分析和实验结果表明,该算法具有较高的估计精度,与一般的粒子滤波算法相比,有较好的稳定性和较低的计算量.

Particle Filter Algorithm Based on Statistical Linear Regression

In this paper, a new particle filter based on Statistical Linear Regression (SLR) is proposed for the state estimation of non-Gauss nonlinear systems. In the new algorithm, the
importance density function of particle filter is generated by linearizing the nonlinear function using statistical linear regression through a set of Gauss-Hermite quadrature points estimating regression coefficient. The density function integrates the new observations into system state transition and extends the overlap fields with true posterior density. The simulation shows that the new algorithm not only has high estimation accuracy but also has better stability and less computation amount than the PF.