基于情景萤火虫算法的机器人路径规划

发布时间：2019-01-01 08:11

【摘要】：针对传统非结构环境下路径规划时机器人运动状态振荡和适应性差等问题,提出了一种基于情景萤火虫算法(SGSO)的机器人路径规划策略。该算法基于混沌系统的规律性、随机性和历遍性以实现初始化,并利用黄金比分割法进行后期优化,以提高种群的多样性,抑制算法的早熟和局部收敛;同时,引入关于萤火虫"天敌"的情景理解,改进萤火虫种群的选择机制,解决萤火虫在非结构环境下寻迹过程中的搁浅现象,增强了算法的适应性和鲁棒性。四个测试函数的仿真实验结果表明,所提算法的求解精度、收敛效率优于基本萤火虫种群优化(GSO)算法;将该算法应用于非结构环境下移动机器人的路径规划中,检测结果表明,基于SGSO的规划路径更短,且转角处更光滑,有效避免了机器人大角度转向对动力系统造成的额外负荷,验证了所提算法的可行性和有效性。
[Abstract]:In order to solve the problems of robot motion state oscillation and poor adaptability in traditional path planning under unstructured environment, a robot path planning strategy based on scenario firefly algorithm (SGSO) is proposed. The algorithm is based on the regularity, randomness and history of chaotic system to realize initialization, and the gold ratio segmentation method is used to optimize the later stage to improve the diversity of population and restrain the precocity and local convergence of the algorithm. At the same time, the scenario understanding of the natural enemy of the firefly is introduced to improve the selection mechanism of the firefly population, to solve the stranded phenomenon in the tracing process of the firefly in the non-structural environment, and to enhance the adaptability and robustness of the algorithm. The simulation results of four test functions show that the proposed algorithm has better accuracy and convergence efficiency than the basic firefly population optimization (GSO) algorithm. The algorithm is applied to the path planning of mobile robot in unstructured environment. The detection results show that the path planning based on SGSO is shorter and the corner is smoother, which effectively avoids the extra load caused by large angle steering of the robot to the power system. The feasibility and effectiveness of the proposed algorithm are verified.
【作者单位】：重庆交通大学机电与车辆工程学院;
【基金】：国家自然科学基金资助项目(51375519)~~
【分类号】：TP242

【相似文献】