Abstract:

Aiming at the particle degradation problem of an mobile robot FastSLAM  a chaos optimization MPSO based algorithm was proposed. The algorithm incorporated the newest observation information into the prediction of particle, adjusted the proposal distribution of the particles, and the accuracy of prediction of a robot's position was enhanced. The MPSO was solved by a sequential two-step optimization strategy. Firstly, the speed of evolution of particle was improved by the median-oriented acceleration, the particle degradation effectively was overcome, the convergence of the algorithm was improved. Then, focusing on the depletion of the particle, the chaos search algorithm optimization algorithms was introduced to MPSO global optimal position to disperse gathered at local optimum particle swarm to the global optimum location close to broaden the scope of the solution space, thus maintaining the population the diversity of simulation. The experimental results prove that the improved method is correct and feasible.

Key words: fast simultaneous location and mapping(FastSLAM), proposal distribution, media-oriented particle swarm optimization(MPSO), median-oriented acceleration, chaos

摘要：

针对移动机器人快速同时定位和地图创建(FastSLAM)中粒子退化问题，提出一种基于混沌优化的中值导向粒子群优化(MPSO)算法。该算法在粒子估计过程中引入观测信息，调整粒子的提议分布，提高位置预测的准测性。混沌优化MPSO算法采用两步优化策略，首先通过中值导向加速度来改进粒子的进化速度，有效地克服粒子退化问题，改善算法的收敛性;然后针对粒子耗尽问题，在MPSO优化算法中引入混沌搜索算法来寻找全局最优位置，驱散聚集在局部最优的粒子群，使其向全局最优位置靠近，扩大解空间的范围，从而保持种群的多样性。仿真和实时数据证明了该方法正确、可行。

关键词: 快速同时定位和地图创建, 提议分布, 中值导向粒子群优化, 中值导向加速度, 混沌