基于高维多目标多约束分组优化的要地防空扇形优化部署

Key-point air defense fan-shaped deployment with large-dimensional multi-objective multi-constraint group divided optimization

基于排队论,得出了信息熵形式的防空阵地网对要地保护能力的数学描述,并将防空阵地网对要地保护能力、防空武器系统部署地点地形条件和防线划分等因素作为优化目标,防空武器系统最小部署间距为约束条件,建立了多型防空武器扇形优化部署多目标优化模型。针对目前算法在解决高维多目标优化存在的问题,基于改进的强度帕累托进化算法(strength Pareto evolutionary algorithm,SPEA2),提出了将多个目标函数分成若干组,分别寻优,再综合求取全体目标函数非支配集的分组优化算法。仿真实验证明,该优化模型能够按现代防空作战特点进行防空武器系统的防线部署,规避不良地形,形成严密的防空覆盖面,且分组优化算法在性能上优于当前高维多目标优化降维算法。

The air defense line intercept ability in the form of information entropy is introduced based on the queuing theory. An optimization model is established taking into account the defense ability of the air defense network, the influence of terrain conditions on air defense weapons to be deployed and the division of the defensive line, minimum distance of two air defense weapons is regarded as constraints. For the reason that SPEA2 is inefficient in large-dimensional multi-objective optimization, a group divided dimensional reduction algorithm is employed. For this algorithm, all the objective functions are divided into some groups, the sub non-dominated set of every function group is built separately, then the non-dominated set of all objective functions is obtained based on the sub non-dominated sets. Simulation results show that using this optimization model can deploy air defense weapons in a proper defensive line where the terrain condition is better, form a tight air defense network, and the performances of group divided dimensional reduction algorithm are better than other large-dimensional Multi-objective optimization.