Research On Simultaneous Localization And Mapping Data Association Algorithm Based On Grey Wolf Optimization Algorithm

Simultaneity localization and mapping are one of the key technologies for mobile robot autonomous movement in an unknown environment.When the robot starts,it can locate its position through continuous observation of the surrounding environmental characteristics,and then build a map of the surrounding environment based on the obtained observation information,to achieve the purpose of map building and positioning.Firstly,aiming at the problem to be studied,the system model of the robot is built to conduct basic learning and understanding of the robot,thus laying a solid foundation for the follow-up research on data association algorithm in SLAM.Secondly,the EKF-SLAM algorithm based on the extended Kalman filter is studied,and the feasibility of EKF-SLAM is verified,and the errors generated by the final system are analyzed.Through analysis,it can be seen that in the process of SLAM,the computational amount and dimension of data association will become larger and larger,which is also the key problem to be solved in this paper.With the help of the idea of intelligent optimization algorithm,three improvement strategies are proposed to improve the efficiency of the grey Wolf optimization algorithm,which is easy to fall into the local best.Logistic chaos map initialization algorithm population strategy: the improved Logistic chaos map is added to initialize the population of the algorithm,which can make the individuals in the population better distributed in the solution space,so as to improve the search efficiency of GWO algorithm;Distance control parameters improvement strategies: the distance will decrease the linear control parameters change for nonlinear decreasing distance control,in the early stage of the iterations,the larger distance control parameters can improve the searching ability of algorithms,late in the iteration,the smaller the distance of the control parameters can be around to ensure that the algorithm in the current solution development ability;Random frog jump worst position change strategy: Worst position change strategy introduced stochastic leapfrog algorithm can effectively overcome the population into a local optimum and the shortcoming of stagnation,through the simulation,the algorithm of SFL-GWO comes under the condition of the high dimension and large amount of calculation,still can spend little time to find the optimal solution,through the comparison and analysis of multiple algorithm show that SFL-GWO algorithm improvement is effective.Again,the data association of SLAM,analyzing relevant theory and model design to achieve the Nearest Neighbor and Joint Compatible Branch and Bound association algorithm and analyzes the shortage of the by converting the data correlation model to combinatorial optimization problems,introduces Mahalanobis distance instead of the traditional Euclidean distance feature,the minimum distance between observations and feature,SFL-GWO algorithm is used to find the minimum distance between all effective observations and the map existing features.Aiming at the shortcomings of the JCBB algorithm in data association,the SFL-GWO is proposed to optimize the JCBB algorithm,which is used to optimize the association algorithm when the number of features is increasing.Improve the accuracy and real-time in the process of data association.Finally,the NN and JCBB association algorithm and based on SFL-GWO JCBB association algorithm combined with EKF-SLAM algorithm were verified by simulation experiments in the designed square corridor environment,and the correlation accuracy,error rate,misjudgment rate,and error result graphs of the three data association algorithms were obtained.Through comparative analysis,it can be seen that the based on SFL-GWO JCBB algorithm can effectively improve the accuracy of data association,and also ensure the real-time performance of the algorithm.