The　ability　to　simultaneous　localization　and　precise　mapping　is　a　predetermination　of　autonomous　robots.　But　because　of　the　unknown　location,　the　unpredictable　environment　information,　the　method　of　simultaneous　localization　and　map　building　based　on　improved　Particle　Filter　in　grid　map　is　presented　to　solve　these　uncertain　problems.　The　distribution　over　robot　poses　and　map　information　is　estimated　with　Bayes＇　rules　and　the　improved　Particle　Filter　respectively.　The　simulation　result　shows　the　method　reduces　the　complexity　of　the　data　and　enhances　the　real-time　performance　of　the　improved　algorithm.　With　the　method,　the　robot　localizes　itself　accurately　as　well　as　builds　grid　map　with　higher　accuracy.　The　proposed　algorithm　is　effective　and　reliable　to　realize　simultaneous　localization　and　mapping.　©　2011　IEEE.