The Research Accurate Position Estimation Of Mobile Robot Based On Improved Particle Filter

In the research of tracking and positioning of mobile robots,Kalman filtering is a commonly used tracking and positioning method.However,this algorithm requires the system model to meet the characteristics of linearity,Gaussian distribution,which is difficult to apply to modern control systems with increasing complexity in practice.However,the particle filter algorithm solves the limitation of the system model limitation,which plays a huge role in the state estimation of nonlinear non-Gaussian systems.Due to the existence of various gross errors in the measurement of mobile robots,system state estimation based on traditional particle filter algorithms will produce large deviations,which will seriously affect the accuracy of mobile robot position estimation.For the position estimation of mobile robot,this paper deeply researches how to improve the traditional particle filter to deal with the gross errors so as to achieve accurate position estimation.The main research contents are as follows.(1)For the problem of inaccurate mobile robot position estimation,its dynamic model is analyzed.The main sensors used in the mobile robot system and its system modeling are introduced,which lays the foundation for experiments.(2)For the problem of gross errors in the system.Firstly,the reasons for the three gross errors of outlier,biases,and systematic drift and their effects on state estimation are explained in detail.Then,the characteristics of three gross errors are analyzed,and methods for detecting,identifying and compensating gross errors are designed.On this basis,an improved particle filter algorithm for accurate position estimation of mobile robots is proposed,and the main ideas and advantages of the algorithm are described.(3)For the problem of mobile robot position estimation with gross errors,a simulation model was used to verify the example.Based on the description of the mobile robot system model,according to the real situation,the types and parameter changes of corresponding measurement errors are generated and simulated.By comparing the parameter changes of position estimation based on the particle filter algorithm and the improved particle filter algorithm,the effectiveness and advantages of the proposed improved particle filter algorithm are verified,and a robust estimation of the system state is achieved.This paper conducts a series of studies on the problem of mobile robot's position estimation.For the problem of gross errors in state estimation,an improved algorithm based on particle filter algorithm is proposed.The improved algorithm solves the problem of inaccurate mobile robot position estimation when three gross errors occur simultaneously in the system,and the effectiveness and advantages of the proposed method are verified through simulation experiments.