Cubature MCL: Mobile robot Monte Carlo Localization based on Cubature Particle Filter

Abstract

Particle Filter is the key issue in mobile robot MCL (Monte Carlo Loclaization, MCL). To overcome particle set degeneracy phenomenon of the traditional MCL algorithm, a new Cubature MCL algorithm is proposed in this paper. The proposed Cubature MCL algorithm utilizes cubature Kalman filter to generate more accuracy proposal distribution, which introduce most recent measurements into Sequential Importance Sampling (SIS) routine of the particle filter. The performance of the Cubature MCL algorithm is presented and analyzed in simulations. The results verify the effectiveness of the proposed Cubature MCL algorithm. The Cubature MCL provides a valuable reference for the mobile robot localization algorithm research.