Abstract:

In a static environment,to solve the problems of particle degradation and reduced particle diversity in the parameter estimation process of the particle filter algorithm,a new algorithm combining the unscented Kalman filter algorithm and improved differential evolution algorithm is proposed to optimize the particle filter.The algorithm first introduces the unscented Kalman filter in the importance sampling stage of the particle filter to calculate the mean and covariance of each particle,and uses the mean and covariance to "guide" sampling to avoid particle degradation.Second,an adaptive strategy is adopted in the mutation and crossover process of traditional differential evolution to avoid premature convergence.At the same time,the improved differential evolution algorithm is used to replace the particle filter's resampling process,which overcomes the problem of reduced particle diversity.Finally,the new algorithm is used to realize multipath parameter estimation.Simulation results show that the new algorithm can effectively reduce the fluctuation range of the parameter estimation results and reduce the root mean square error while meeting the real-time requirements.

Key words: parameter estimation, particle filter, particle degradation, improved differential evolution