为提高4目标以上高维多目标优化问题的求解性能,提出一种基于改进K支配排序的高维多目标进化算法(KS-MODE).该算法针对K支配的支配关系和排序方法进行改进,避免循环支配并增强选择压力;设计新的全局密度估计方法提高局部密度估计精确性;设计新的精英选择策略和适应度值评价函数;采用CAO局部搜索算子加速收敛.在4～30个目标标准测试函数上的实验结果表明,KS-MODE能够在保证解集分布性的同时大幅提升收敛性和稳定性,能够有效求解高维多目标优化问题. In order to improve the performance of solving high-dimensional multi-objective optimization problems with 4-objective, a new high-dimensional multi-objective evolutionary algorithm (KS-MODE) based on improved K-dominated ordering is proposed.This algorithm improves on the dominance relation and sorting method of K- Avoid circular control and enhance the selection pressure; design a new global density estimation method to improve the accuracy of local density estimation; design a new elite selection strategy and fitness value evaluation function; use CAO local search operator to accelerate convergence in 4 to 30 targets Experimental results on standard test functions show that KS-MODE can significantly improve the convergence and stability while guaranteeing the distribution of solution sets, and can effectively solve the high-dimensional multi-objective optimization problem.