Knowledge-Guided Evolutionary Optimization for Large-Scale Air Defense Resource Allocation

In the high-stakes domain of air defense, resource allocation is a challenging optimization problem that involves many constraints and requires fast solutions. Previous work mainly addressed the single-intercept sensor-weapon-target assignment problem, which assigns missiles and radars to targets without optimizing the launch time. However, this problem can become large-scale and difficult to solve when saturation attack tactics are employed. Therefore, we propose a more realistic model that allows multiple missiles to intercept the same target and considers the launch time as a decision variable. We also incorporate radar relay guidance to maximize the interception probability and minimize the missile cost. To solve this model efficiently, we present a novel evolutionary algorithm with linear population size reduction (LPSR) and knowledge-guided strategies, called LKEA. Specifically, we design a heuristic method to quickly find the optimal launch time for each missile and a knowledge-guided strategy to use the radar channel collision information to guide the evolution. Experimental results show that LKEA is competitive in solving small/medium/large air defense resource allocation problems (ADRAPs).