基于膜计算粒子群优化的FastSLAM算法改进

Improved FastSLAM Algorithm Based on Particle Swarm Optimization of Membrane Computation

针对FastSLAM算法存在的粒子退化和粒子多样性缺失问题,提出了一种基于膜计算粒子群优化的FastSLAM算法.该算法将膜计算和粒子群优化算法相结合,利用膜计算的并行性、分布式的特点和粒子群优化算法的简单高效的优点,加速调整FastSLAM算法中粒子群的建议分布向全局最优解处收敛,在保证算法局部搜索精度的同时,扩大搜索范围,提高全局搜索的多样性,促使预测粒子更快的朝着真实的机器人位姿状态逼近,减缓粒子退化.最后利用MATLAB平台进行仿真实验.实验结果表明该算法提高了FastSLAM算法的定位精度,同时减少了系统运行时间,效率得到有效提高.

In order to solve the problem of particle degradation and lack of diversity in traditional FastSLAM algorithm,a FastSLAM algorithm based on Particle Swarm Optimization for membrane computing was proposed.This algorithm combines membrane computing with particle swarm optimization(PSO).By taking advantage of the strong parallelism and distributed characteristics of membrane computing and the simple and efficient advantages of PSO,the proposed distribution of particle swarm in FastSLAM algorithm can be adjusted to converge to the global optimal solution,and the local search accuracy of the algorithm can be guaranteed while enlarging.The search range improves the diversity of global search,promotes the predicted particles to approach the real robot position and posture faster,and slows down the particle degradation.Finally,the simulation experiment is carried out on the platform of MATLAB.The experimental results show that the algorithm improves the positioning accuracy of FastSLAM algorithm,reduces the running time of the system,and improves the efficiency effectively.