基于改进灰狼算法的无人矿车路径规划

Path Planning of Unmanned Mine Truck Based on Improved Grey Wolf Algorithm

针对露天矿山复杂地形环境下常规的运输路径规划方法易陷入局部最优、路径收敛速度慢、耗时长等问题,提出了一种改进的灰狼算法,对无人矿车进行路径规划。考虑矿区地形坡度起伏变化的特点,通过建立坡度-速度模型,将矿区栅格环境中矿车的速度引入到上下坡行驶状态转移规则中;构造一种结合正余弦变换的收敛因子,以更好地平衡全局和局部的搜索能力;将最优灰狼个体进行随机交叉扰动,加强狼群跳出局部最优的能力。仿真试验结果表明,在20 km×20 km复杂的矿区栅格环境下,改进的灰狼算法在规划平均路径和平均运行时耗上较遗传算法、正余弦算法和传统灰狼算法分别缩短了2.53%~4.77%和3.97%~12.32%。在寻优能力、收敛速度以及稳定性等方面均优于传统算法,研究可为智慧矿山建设中露天矿山无人矿车的运输作业工作提供借鉴。

In view of the shortcomings of conventional transportation path planning methods in complex terrain environments of open-pit mines,such as being prone to local optima,slow path convergence speed,and long time consumption,an improved grey wolf algorithm was proposed for path planning of unmanned mine truck.Considering the characteristics of terrain slope fluctuations in mining areas,a slope-speed model was established to introduce the speed of mine truck in the grid environment of the mining area into the transition rules of uphill and downhill driving states.A convergence factor combining sine and cosine transform was constructed to better balance the global and local search ability.The optimal grey wolf individuals were cross perturbed to enhance the ability of the wolves to jump out of the local optimum.The simulation test results show that in a complex grid environment of 20 km×20 km mining area,the improved grey wolf algorithm reduces the average path planning and average running time by 2.53%-4.77%and 3.97%-12.32%compared to genetic algorithm,cosine algorithm,and traditional grey wolf algorithm,respectively.In terms of optimization ability,convergence speed,and stability,it is superior to traditional algorithms.The research can provide a reference for the transportation operation of unmanned mine trucks in open-pit mines in the construction of intelligent mines.