摘要
针对基于灰狼优化算法的机器人在路径寻优过程中易陷入局部最优和后期收敛速率慢的问题,提出一种混合粒子群与改进灰狼的算法(PGWO)。通过粒子群算法确定狼群初始适应度值,改进传统灰狼中的收敛因子,在平衡其搜索能力的同时,提高算法后期的搜索速率,对种群权重进行动态分配来降低陷入局部最优的可能性,将粒子群与改进的灰狼算法混合运行后获得最优解。结果表明,在两种栅格地图上,PGWO与传统灰狼算法相比,路径分别缩短了35.03%和34.58%,搜索时间分别减少了52.69%和51.06%,收敛速度分别提升了30.62%和34.30%;与改进灰狼算法相比,路径分别缩短了22.03%和23.04%,搜索时间分别减少了33.05%和25.81%,收敛速度分别提升了16.83%和20.98%;与蚁群算法相比,路径分别缩短了24.08%和25.41%,搜索时间分别减少了65.04%和79.74%,收敛速度分别提升了23.53%和32.34%,说明了PGWO算法在路径寻优中的有效性。
Aiming at addressing the challenges of local optima trapping and slow convergence rates encountered by robots in path planning based on the grey wolf optimization algorithm,this paper proposes a hybrid algorithm called particle swarm optimization and improved grey wolf optimization(PGWO).The PGWO algorithm leverages particle swarm optimization to determine the initial fitness values of the wolf pack and enhances the convergence factor in traditional grey wolf optimization,thereby balancing its search capability and improving the algorithm's late-stage search rate.Additionally,dynamic allocation of population weights is employed to reduce the likelihood of falling into local optima.By blending particle swarm optimization with the improved grey wolf algorithm,the optimal solution is attained.Results demonstrate that,compared to traditional grey wolf algorithm,PGWO reduces the path length by 35.03%and 34.58%,decreases search time by 52.69%and 51.06%,and improves convergence speed by 30.62%and 34.30%on two grid maps,respectively.Compared to the improved grey wolf algorithm,PGWO reduces the path length by 22.03%and 23.04%,decreases search time by 33.05%and 25.81%,and improves convergence speed by 16.83%and 20.98%,respectively.Furthermore,compared to ant colony optimization,PGWO shortens the path length by 24.08%and 25.41%,reduces search time by 65.04%and 79.74%,and improves convergence speed by 23.53%and 32.34%,indicating the effectiveness of the PGWO algorithm in path planning optimization.
作者
莫定界
赵杰
凌港
张冬青
陈嘉晋
MO Dingjie;ZHAO Jie;LING Gang;ZHANG Dongqing;CHEN Jiajin(School of Electrical&Control Engineering,Heilongjiang University of Science&Technology,Harbin 150022,China)
出处
《软件导刊》
2025年第7期54-60,共7页
Software Guide
基金
黑龙江省省属高等学校基本科研业务费项目(2022-KYYWF-0551)。
