巷道断面成形轨迹的多目标蚁群规划方法  

Trajectory planning method based on multi-objective ant colony algorithm for forming section of roadway

作　　者：郭一楠 黄遥 巩敦卫 夏煌煌 GUO Yinan;HUANG Yao;GONG Dunwei;XIA Huanghuang(School of Mechanical Electronic and Information Engineering,China University of Mining and Technology-Beijing,Beijing 100083,China;School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221116,Jiangsu China)

机构地区：[1]中国矿业大学(北京)机电与信息工程学院,北京100083 [2]中国矿业大学信息与控制工程学院,江苏徐州221116

出　　处：《华中科技大学学报（自然科学版）》2022年第5期115-122,共8页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61973305,52121003);高等学校学科创新引智计划资助项目(B21014).

摘　　要：针对现有的巷道断面轨迹规划方法优化目标单一且忽略巷道煤岩分布特征的问题,综合考虑最佳成形质量和最低截割能耗,以巷道断面煤岩分布和掘进设备特性作为约束,构建截割轨迹多目标规划模型,提出一种改进的多目标蚁群算法,以实现模型求解.在算法中,为了避免随机初始化成形轨迹而导致的搜索效率低且轨迹不可行程度高等问题,抽取轨迹自身先验知识,设计初始信息素不均匀分配策略;提出一种基于目标重要性系数的自适应调节策略,防止搜索陷入局部最优和搜索停滞.基于实际巷道断面的夹矸分布的复杂性,设置了四种巷道断面仿真环境,对所提策略中参数的重要性及策略自身的有效性进行验证,结果表明:与其他四种算法相比,所提算法具有较好的收敛性和分布性,在巷道断面成形轨迹规划问题中可以获得较优的解.Aiming at the problem that the single optimized objective was normally taken and the distribution of coal and rock was ignored in existing trajectory planning methods for forming section of a roadway,a multi-objective trajectory planning model was built by taking the coal and rock distribution of roadway section and the characteristics of tunneling equipment as constraints,in which the best forming quality and the lowest energy-consumption of excavating were comprehensive considered,and an improved multi-objective ant colony algorithm was proposed to solve the problem.In the proposed algorithm,to avoid low search efficiency and high infeasibility caused by random initialized trajectory,the prior knowledge of paths was extracted and the imbalanced assigning strategy of the initial pheromone was designed,and an adaptive adjustment strategy for the objective importance coefficient was proposed to prevent the search from falling into local optimal and search stagnation.Considering the complexity of the gangue distribution in actual roadway sections,four simulation environments of roadway sections were constructed to verify the sensitivity of parameter and the effectiveness of the proposed strategies.Results show that compared with the other four algorithms,the proposed algorithm has better convergence and distribution,and the optimal trajectory can be obtained in the trajectory planning for forming section of a roadway.

关 键 词：轨迹规划 多目标优化 改进多目标蚁群算法 截割能耗 欠挖面积 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]