针对复杂环境下的无人机多航迹规划问题,提出了将粒子群优化(particle swarm optimization,PSO)算法与加权k-均值聚类算法相结合的规划方法。每个粒子表示一条航迹,采用加权k-均值聚类算法对粒子进行分类,得到多个粒子子群,在每个子群内部进行一条可行航迹的优化,最终得到多条不同的可行航迹。对传统k-均值聚类算法进行改进,采用排挤机制产生初始聚类中心,针对实际环境中突发威胁的分布不均性,在聚类过程中,对航迹节点按照所在区域突发威胁的出现概率进行加权,提出了加权k-均值聚类算法。仿真实验表明,所提出的方法能够有效地得到无人机的多条可行航迹。 In order to solve the multi-track planning problem of UAV in complex environment, a planning method combining particle swarm optimization (PSO) with weighted k-means clustering algorithm is proposed. Each particle represents a track, and the weighted k-means clustering algorithm is used to classify the particles to obtain multiple particle subgroups. One feasible track is optimized within each subgroup to finally obtain a plurality of different feasible tracks . The traditional k-means clustering algorithm is improved, and the crowding-out mechanism is used to generate the initial cluster centers. According to the uneven distribution of sudden threats in the real environment, in the clustering process, according to the sudden threat The probability of occurrence is weighted, a weighted k-means clustering algorithm is proposed. The simulation results show that the proposed method can effectively obtain a number of feasible trajectories of the UAV.