摘要
首先根据仓储物流环境的特点构建了可灵活扩展的仓储环境模型，并制定了适合仓储物流需求的机器人运动规则，使该模型能够适用于动态的仓储物流环境；其次采用线性时序逻辑任务公式描述具体的任务需求，使其可以适用于实际应用中更加复杂的任务；继而将任务需求与环境信息相融合，构建任务可行网络拓扑，避免分段任务搜索；然后采用Dijkstra算法在任务可行网络拓扑上搜索出最优路径，确保规划所得路径的最优性；最后将任务可行网络拓扑上的最优路径映射回加权切换系统，获得环境中满足任务需求的最优路径。与目前广泛使用的A＊算法相比，上述方法不仅能够满足复杂的任务需求，而且能够保证路径规划的最优性，而不是次优解。
The path planning for warehouse robots based on the linear temporal logic （LTL） theory was studied. Firstly, an extensible warehouse model was built up according to warehouse environmental characteristics, and a set of rules for governing robot movements were defined to make the model adapt dynamic warehouse environments. Secondly, the LTL formula was used to describe task requirements to make it suitable for more complicated tasks in practical applications. Thirdly, the task-feasible network topology was built by combining task requirements and environmen- tal information to avoid piecewise path searching. Then, the Dijkstra algorithm was utilized to search the global op- timal path on the task-feasible network topology to ensure the optimality of the path planning result. Finally, the optimized path was obtained by mapping the optimal path on the task feasible network topology back to the ware- house model. The above-mentioned method can not only satisfy complex task requirements, but also can ensure the optimality of the path planning result, rather than a suboptimal solution, compared with the current widely used A ＊ algorithm
出处
《高技术通讯》
CAS
CSCD
北大核心
2016年第1期16-23,共8页
Chinese High Technology Letters
基金
国家自然科学基金(61273116)
863计划(2014AA041601-05)
浙江省自然科学基金(LY15F030015)资助项目
