基于无人叉车行驶时间优化的改进 A $ ^{*} $ 算法

传统 A $ ^{*} $ 算法规划所得路径拐点个数多、转弯角度大,导致无人叉车在行驶过程中需要频繁进行制动、转向等操作,增加了行驶时间和危险系数。为解决上述问题,文章以无人叉车为研究对象,采用8向栅格地图对环境建模,提出一种基于行驶时间优化的 A $ ^{*} $ 算法。通过引入转弯行驶时间及转弯代价补偿参数优化实际代价函数,以减少路径中的拐点;向启发函数引入父节点启发函数,同时对启发函数的权重进行自适应调整,使搜索方向在更快地向目标节点靠近的同时保证路径最优。仿真实验结果表明:改进 A $ ^{*} $ 算法有效且实用;相较于传统 A $ ^{*} $ 算法,改进 A $ ^{*} $ 算法在 $ 30 \times 30 $ 栅格地图中拐点个数减少 40.00%,转弯总角度减少 60.00%;在 $ 50 \times 50 $ 栅格地图中拐点个数减少 47.10%,转弯总角度减少 65.90%。

The traditional A* algorithm plans a path with many inflection points and large turning angles, which leads to frequent braking and steering operations of unmanned forklifts during the driving process, increasing the driving time and danger factor. To solve the above problems, this paper takes unmanned forklift as the research object, adopts eight-way grid map to model the environment, and proposes an A* algorithm based on driving time optimization. The actual cost function is optimized by introducing a turning driving time function and turning cost compensation parameter to reduce the inflection points; the parent node heuristic function is introduced to the heuristic function, while the weights of the heuristic function are adaptively adjusted to make the search direction approach the target node faster while ensuring the optimal path. The effectiveness and practicability of the improved A* algorithm are verified by the simulation test. Comparing the simulation results of the improved A* algorithm and the traditional A* algorithm, the number of inflection points is reduced by 40.00% and the total turning angle is reduced by 60.00% in the $ 30 \times 30 $ grid map; the number of inflection points is reduced by 47.10% and the total turning angle is reduced by 65.90% in the $ 50 \times 50 $ grid map.