基于改进ACO-GA算法的臂架布料避障规划研究  

Obstacle Avoidance Path Planning for Concrete Placing Booms Using an Improved ACO-GA Fusion Algorithm

作　　者：吴辉 WU Hui(Huxin Cement CO.,Ltd.,Wuhan 430205,China)

机构地区：[1]华新水泥股份有限公司,湖北武汉430205

出　　处：《数字制造科学》2024年第4期284-289,共6页

摘　　要：针对混凝土布料臂架在浇筑平面时,采用传统蚁群优化算法(ACO)进行避障路径规划,其仅以避障路径代价为路径评价指标,规划路径通常非最优解等问题,提出融合改进的ACO-GA避障路径规划算法。通过引入信息素矩阵调整策略并改进启发函数,提升ACO算法搜索性能与收敛速度。针对ACO规划的路径非全局最优、平滑性差等问题,引入了遗传算法(GA),将初步规划的路径集合作为GA算法的初始种群,提出考虑路径平滑性与控制节点数量的适应度函数与适应度加权交叉算子,使得规划的路径质量更优。仿真场景与模拟避障场景试验结果表明,提出的改进ACO-GA算法在规划避障路径时,路径长度、平滑性与算法收敛速度等指标均优于改进前算法,且规划路径符合臂架的运动要求。When the traditional Ant Colony Optimization(ACO)algorithm is applied to obstacle avoidance path planning for concrete placing booms during pouring,it often results in suboptimal paths,as it relies solely on obstacle avoidance path cost as the evaluation criterion.To address this,an improved ACO-GA fusion algorithm is proposed.The approach enhances the search performance and convergence speed of the ACO algorithm by introducing a pheromone matrix adjustment strategy and an improved heuristic function.To tackle the issues of non-global optimality and poor smoothness in ACO-generated paths,the Genetic Algorithm(GA)is integrated,using the initial path set generated by ACO as the initial population for GA.A custom fitness function is designed,incorporating considerations for path smoothness and the number of control nodes,along with a fitness-weighted crossover operator to further enhance path quality.Simulation scenarios and experimental results in obstacle avoidance environments demonstrate that the improved ACO-GA algorithm surpasses the original algorithms in terms of path length,smoothness,and algorithm convergence speed.Additionally,the planned paths meet the motion requirements of the placing boom system.

关 键 词：改进蚁群算法 改进遗传算法 混凝土布料臂 避障规划 

分 类 号：TU646[建筑科学—建筑技术科学] TP27[自动化与计算机技术—检测技术与自动化装置]