【摘要】 针对无人扫地车工作时路径重复率高，转弯次数多影响移动速度而造成额外能耗和时间的问题，提出了一种改进单元分解法和蚁群算法相结合的全覆盖路径规划方法。首先，在单元分解时考虑相邻区域规划将凹型区域有选择性的分解；其次，考虑到路径规划的转弯耗能，在子区域内往返遍历时引入一种转弯成本代价函数；然后，针对传统蚁群算法提出改进初始信息素为非均匀分布和自适应调整启发式函数来加快收敛速度，用改进信息素更新方法将最差蚂蚁对算法性能的影响降至最低，以避免陷入局部最优解；最后，用改进蚁群算法对各个子区域进行单元间的规划衔接来实现全覆盖路径规划。仿真结果表明改进后的全覆盖路径规划方法可以减少重复率、时间和转弯次数，缩短路径长度和能耗。更多还原

【Abstract】 Aiming at the problem that the high path repetition rate and the large number of turns of the autonomous floor-sweeping vehicle during operation affect the moving speed, resulting in additional energy consumption and time, a full-coverage path planning method combining an improved cell decomposition method and an ant colony algorithm is proposed. First, during the unit decomposition process, neighboring areas are considered for selective decomposition of concave areas.Second, considering the energy consumption of turning in path planning, a turning cost function is introduced during the round-trip traversal in sub-regions. Then, an improved initial pheromone distribution and adaptive adjustment of heuristic function are proposed for the traditional ant colony algorithm to accelerate convergence speed. The improved pheromone update method minimizes the impact of the worst ant on the algorithm performance to avoid falling into local optimal solutions. Finally, using the improved ant colony algorithm to perform inter-unit planning between sub-regions to achieve full coverage path planning. The simulation results show that the improved full coverage path planning method can reduce repetition rate, time and turning times, shorten path length and energy consumption.更多还原