Multi-weapon multi-target assignment based on hybrid genetic algorithm in uncertain environment
Zhao Yang, Yifei Chen, Ziyang Zhen, Ju Jiang
Abstract
The multi-weapon multi-target assignment is always an unavoidable problem in military field. It does make sense to find a proper assignment of weapons to targets which may help maximize the attack effect. In this article, as the information achieved from the battlefield is becoming more and more uncertain, a novel threat assessment method and target assignment algorithm are proposed against the background of unmanned aerial vehicles intelligent air combat. Specifically, with regard to the threat assessment issue, a possibility degree function based on grey theory is structured to further improve the grey analytic hierarchy process. It can transform the interval weight of threat factors into scalar-valued weight, with which the accuracy of threat assessment can be improved. Regarding the target assignment problem, combining with interval grey number, an improved hybrid genetic algorithm is developed. The improvements are mainly consisting of adaptive crossover and mutation operators which can help to find an approximate solution within certain time constraints. Meanwhile, the simulated annealing operation is incorporated to avoid local optimum and premature phenomenon. In addition, the selection operation and fitness function are also redesigned to handle the interval numbers. Simulation results demonstrate the effectiveness of our algorithm in completing the multi-objective weapon-target assignment under uncertain environment.