论文部分内容阅读
针对一类移动机器人航迹推算系统的故障诊断问题,提出一种多模态进化Rao-Blackwellized粒子滤波器(MERBPF)算法.为解决由粒子贫乏引起的不一致性问题,采用交叉与变异种群策略优化,根据粒子多样性加入扰动因子.利用专家规则判定机器人运动状态所对应的MERBPF,构造复杂逻辑表述方法.仿真实验结果表明:在强过程噪声下,MERBPF仍具有较高的鲁棒性,提高了诊断机器人航迹推算系统的准确率.
In order to solve the problem of inconsistency caused by particle depletion, a multi-modal Evolutionary Rao-Blackwellized Particle Filter (MERBPF) algorithm is proposed to solve the fault diagnosis of a class of mobile robot trajectory estimation system. , Adding perturbation factor according to the particle diversity.Make use of expert rules to determine the MERBPF corresponding to the robot’s motion state to construct complex logic representation method.The simulation results show that MERBPF still has high robustness under strong process noise and improves Diagnosis robot trajectory estimation system accuracy.