Research Of Multi-robot SLAM Methods Based On Bayesian

In the field of mobile robot navigation, mobile robot autonomous localization is the mostimportant research direction in the realization of robot truly autonomous and intelligent. Withthe advancement of technology and the urgent needs of people’s lives, simultaneous localizationand mapping (SLAM) problem of multi-robot under an unknown environment has become aresearchhotspot.Themainarticle readsas follows:1. Expounds SLAM methods based on bayesian filtering probability theory, SLAMproblem to be understood as bayesian filtering problem. SLAM bayesian filtering processincludes two steps: movement(time)updated and observations update,establishes multi-robotmotion model and sensor observation model, describes basic theory of multi-robot SLAMmethod.2. Research three algorithms for solving the single robot SLAM problem. A detailedanalysis of extended kalman filter SLAM method, the extended information filtering SLAMmethod and exactly sparse information filtering SLAM method, they are based on Bayesiantheory. In a single robot application, it analyzes their respective characteristics throughsimulationexperiments.3. Research multi-robot SLAM problem. Methods solving a single robot SLAM problemextended to resolve multi-robot SLAM problem. According to bayesian probability theory,method of multi-robot SLAM based on extended information filtering algorithm is deducedformulas and charts analysis, then establishes a complete multi-robot SLAM problemsimulationsystemandsimulation.4. Explore a fast algorithm for multi-robot SLAM method. According to the needs of awide range of environmental applications, the article discusses a fast information filteringalgorithm: multi-robot SLAM method based on exactly sparse extended information filtering,Using reasonable sparse strategy, to maintain movement updates and measurement update stagetime a constant, reflects the effectiveness and accuracy of the algorithm, the advantage having a highcomputationalefficiency.