论文部分内容阅读
A large sample size is required for Monte Carlo localization (MCL) in multi-robot dynamic environment, because of the "kidnapped robot" phenomenon, which will locate most of the samples in the regions with small value of desired posterior density. For this problem the crossover and mutation operators in evolutionary computation are introduced into MCL to make samples move towards the regions where the desired posterior density is large, so that the sample set can represent the density better. The proposed method is termed genetic Monte Carlo localization (GMCL). Application in robot soccer system shows that GMCL can considerably reduce the required number of samples, and is more precise and robust in dynamic environment.