Sensor-Weapon-Target Assignment Based on Hopfield Neural Network

This study focuses on the dual-objective optimization problem of Sensor-Weapon-Target Assignment (SWTA), aiming to minimize the expected threat values of incoming targets during combat while considering the minimal consumption of resources for effective resource allocation. We propose an innovative solution based on the Hopfield neural network by transforming the optimization model of the problem and introducing the energy function of the Hopfield neural network, which linearly combines the two optimization objectives. The states of the neural network correspond to the final allocation matrix, and by randomly adjusting weights, the neural network can converge to different solutions, ultimately obtaining the Pareto front. We designed scenarios for both small and large-scale cases in various combat situations and compared this method with traditional algorithms GA and PSO. The comparison algorithms also use linearly weighted objectives to solve the Pareto front. Experimental results indicate that in most cases, the Hopfield neural network demonstrates superior performance, achieving high-quality solutions at a lower time cost. Compared to traditional GA and PSO algorithms, it exhibits outstanding effectiveness in solving the SWTA problem. This research provides an advanced and reliable approach for military combat decision-making and offers valuable insights for the future development of intelligent systems.