基于ε等级约束差分进化的多径估计算法

本文针对基于扩展Kalman滤波（EKF）的多径估计算法需要对非线性观测方程进行线性化.对初值比较敏感，造成估计性能下降的问题，提出了基于智能优化的多径估计算法.该算法将估计误差的二阶矩作为目标函数，将瞬时误差作为约束条件，同时考虑多径参数的先验信息，实现了将多径估计问题转化为具有约束条件的优化问题.然后，利用一种智能优化算法来解决该优化问题.本文采用了ε等级约束差分进化（εCRDE）算法来解决有约束条件的优化问题，并对该算法进行改进，使改进后的εCRDE算法可以实现多径参数的迭代估计.仿真结果表明，与EKF算法相比，在单一多径和2路多径情况下，基于改进εCRDE的多径估计算法都具有更好的估计性能.

Multipath Estimation Algorithm Using ε Constrained Rank-Based Differential Evolution

The observation equation has to be linearized for the multipath estimation algorithm based on Extended Kalman Filter (EKF).To tackle the problem of being sensitive to initial state,which leads to a performance degradation in terms of estimation accuracy,a new multipath estimation algorithm based on intelligent optimization is proposed.Through minimizing the second moment of the estimation error the multipath estimation problem is transferred into an optimization problem with constrained conditions.Furthermore,the instantaneous error is considered as a constrained condition as well as the prior information of the multipath parameters.Then,an intelligent optimization algorithm is used to solve the presented optimization problem.Especially,the ε Constrained Rank-based Differential Evolution(εCRDE) algorithm is adopted.In addition,the εCRDE algorithm is improved to cater for the need of iteration for multipath estimation.Simulation results show that the proposed algorithm outperforms EKF for multipath estimation in the case of single multipath and two multipaths.