On WSN-aided simultaneous localization and mapping based on particle filtering

Abstract

A novel WSN-aided SLAM(simultaneous localization and mapping)method is proposed for mobile robot to solve two main problems in traditional SLAM methods,i.e.,high dimensions in problem spaces and difficulties in multi- target data association.Model for the WSN-aided SLAM method is.built,and noises are analyzed.Then a distributed particle filtering(PF)data fusion algorithm suitable for this method is developed.The key steps,such as particle initializa- tion,prediction,sequential importance sampling,resampling,are particularly analyzed,and the validity and efficiency of the method are testified by simulation experiment.The experiment results demonstrate that,when the PF algorithm is used and the WSN is integrated for aided navigation,the dimensions of problem spaces can be greatly reduced,the multi-target data association problems can be solved and blind nodes can be located with high precision independent of the anchor node.The precision of localization and mapping for mobile robots can be effectively improved,and the error accumulation of inertial navigation can be effectively suppressed especially when the robot closed-loop track is not required.