【摘要】 针对智慧停车场自动导引车(Automated Guided Vehicle, AGV)存取车辆过程中，由于转弯和拥堵导致停车效率低的问题，提出一种基于动态路段代价的停车引导路径规划算法。首先，提出最优方向的概念，考虑终点与起点的相对位置，设置优先扩展位于最优方向位置节点的策略，能够预判节点搜索方向，减少路径转弯次数并节约计算量；其次，设置一种根据场内AGV运行情况动态调整地图路段代价的函数用以准确反映场内已占用道路信息，引入启发式思想和时间窗原理，算法在生成路径时避开代价较高的易拥堵路段。仿真结果表明，上述算法在求解高质量路径的同时，表现出优秀的计算性能。算法可以满足路径的实时计算要求，适用于AGV代客泊车服务。更多还原

【Abstract】 Aiming at the problem of low parking efficiency caused by turns and congestion in the process of AGV accessing vehicles in smart parking lot, the path planning algorithm based on dynamic road segment cost was proposed. Firstly, the concept of optimal direction was defined. Considering the position of the end-point relative to the start-point, the strategy of preferentially expanding the nodes in the optimal direction was set, which can prejudge the node search direction, reduce the number of turns and save the amount of calculation; Then, a function to dynamically adjust the cost of map sections according to the AGV conditions was set, which can accurately reflect the information of the occupied roads, heuristic idea and time window principle were introduced to avoid the congested roads with high cost when generating paths. The simulation shows that the algorithm can solve high-quality paths while saving operation time, showing higher computational performance. The algorithm can meet the real-time calculation requirements of parking path and is suitable for AGV valet parking service.更多还原