多策略蚁群算法求解越野路径规划

针对车辆的越野路径规划问题,设计了以最少行驶时间为目标的多策略蚁群算法。首先,分析了地形坡度和地表属性对于车辆路径规划的综合影响,通过叠加坡度与粗糙度约束建立了禁忌表;其次,一方面引入了自适应调整策略以提高路径搜索的有效性,另一方面设计了双向搜索策略以增加蚂蚁之间的协作能力和成功路径的搜索机率;另外,还提出了子路径多段交叉策略以提高算法的全局搜索能力和收敛速度,在详细叙述改进算法的步骤之后,优化了算法的部分参数取值;最后,就基本算法和改进算法的性能指标、收敛代数和仿真结果进行了比较与分析。实验结果表明,改进算法能够快速有效地实现越野路径规划,较之基本算法有一定的优越性。

Multi-strategy ant colony algorithm for cross-country path planning

According to the vehicle cross-country path planning problem, a multi-strategy ant colony algorithm with the minimum traveling time as the goal was designed. First of all, the synthesis influence of the terrain slope and surface properties was analyzed for vehicle path planning and the table Tabu constructed by stacking constraints of slope and roughness. Secondly, on the one hand, the adaptive adjustment strategy is imported to improve the effectiveness of the searching paths. On the other hand, the bidirectional search strategy was designed in order to increase ants′ collaboration and the probability of the searching successful paths. In addition, the sub-paths multi-segment crossover strategy was also proposed to improve the global searching capability and accelerate the convergence speed. The improved algorithm in detail steps was described, and the value of some parameters are optimized. Finally, the performance indicators, convergence algebra and simulation results between the basic algorithm and the improved algorithm were compared and analyzed. The experimental results show that the improved algorithm can achieve a cross-country path planning in a quick and efficient way and has certain advantages over the basic algorithm.