基于HV-MOPSO的海上通信平台部署多目标粒子群优化研究

Research on Multi-Objective Particle Swarm Optimization for Maritime Communication Platform Deployment Based on HV-MOPSO

摘要: 针对目前海上通信平台部署“提高覆盖率会导致重复覆盖率随之升高、容易产生干扰”的多目标优化问题，文章提出了一种融合超体积(HV)指标的多目标粒子群算法(HV-MOPSO)。该算法通过佳点集实现种群初始化，结合目标空间网格划分维持解集多样性，基于HV指标优化全局最优解选择与外部档案维护，以最小化未覆盖率和最小化重复覆盖率为双目标构建海上通信平台部署优化模型。以WFG系列测试问题为实验对象，与经典MOPSO算法对比，结果表明：HV-MOPSO在多峰欺骗性问题WFG4上的HV均值达81.1845，显著高于对比算法的78.4371，且解集Spacing指标更优，在保持收敛性的同时有效提升解集多样性；将其应用于海上通信平台部署，重复覆盖率降低46%，Spacing指标提升78%。

Abstract: To address the multi-objective optimization problem in the current maritime communication platform deployments—where “an increase in coverage rate leads to a rise in redundant coverage and is prone to interference”—a multi-objective particle swarm optimization algorithm fused with the Hypervolume (HV) indicator (HV-MOPSO) is proposed. This algorithm initializes the population using a good point set, maintains the diversity of the solution set by combining grid division in the objective space, and optimizes the selection of the global optimal solution and the maintenance of the external archive based on the HV indicator. A deployment optimization model for maritime communication platforms is constructed with two objectives: minimizing the uncovered rate and minimizing the redundant coverage rate. Experiments conducted on the WFG series of test problems and a comparison with the classic MOPSO algorithm show that: HV-MOPSO achieves an average HV value of 81.1845 on the multi-modal deceptive problem WFG4, which is significantly higher than the 78.4371 of the comparison algorithm, and its solution set exhibits a superior Spacing indicator, effectively enhancing the diversity of the solution set while maintaining convergence. When applied to the deployment of maritime communication platforms, the redundant coverage rate is reduced by 46%, and the Spacing indicator is improved by 78%.