机器人路径规划是机器人领域的一项重要课题，不同于以往在遗传算法过程中考虑路径平滑度的方法，本文提出了一种将遗传算法过程与路径平滑过程分开的机器人路径规划新方法。先设计可变长编码方式的简单遗传算法产生较优的折线路径，再引入一类新的带形状参数的回旋螺线对其进行平滑操作，以抚平较大转角。整个路径规划过程，只需输入障碍物坐标即可自适应地选择参数以产生机器人行走路径。仿真结果表明，将遗传算法过程与路径平滑过程分离的做法能降低遗传算法本身复杂度，所以设计的平滑操作不仅提高了路径平滑度，还可以减少路径长度。%Robot path planning is an important topic in the field of robotics. Considering the path different from the past in the smoothness of the process of genetic algorithm method, a genetic algorithm process is presented by smoothing the path separate from the new path planning method. After calculating a path using simplified GA method with a variable length coding, a new type of clothoid curve with wipe parameter is introduced to smooth the sharp turns. In the entire path planning process, the parameter can be chosen adaptively to generate a robot’s walk path as soon as the barrier’s coordinate is entered. The simulation results show that the method of separating smooth process from GA process simplifies GA method. The proposed method can efficiently decrease the path length as well as increase the path smoothness.
展开▼
