An estimation of distribution algorithm with efficient constructive repair/improvement operator for the dynamic weapon-target assignment

This paper reports our recent research about new efficient problem-solvers for the dynamic weapon-target assignment (DWTA). A binary-encoding-based estimation of distribution algorithm (EDA) is proposed to solve DWTA problems. An elaborate constructive repair/improvement (CRI) operator is proposed and integrated into the EDA to achieve constraint saturation, which conduces to constraint satisfaction as well as the improvement of generated solutions. The performance comparison against another two EDAs which employ well-known constraint handling methods demonstrates the superiority of the CRI operator. The proposed EDA based on the CRI operator also shows very competitive and even better performance against several state-of-the-art DWTA algorithms.