Research on an Optimization Model for Multi-Objective Smoke Screen Deployment by Drones Based on Genetic Algorithms and Simulated Annealing

Abstract

This paper proposes a multi-layer optimization framework for missile concealment using smoke screens deployed by UAVs. It focuses on coordinating single-UAV single-missile and single-UAV multiple-missile operations through genetic algorithm-based optimization, while incorporating a hybrid solution combining simulated annealing and entropy-weighted evaluation in multi-UAV multiple-missile scenarios. First, three-dimensional dynamic trajectories for missiles, UAVs, warheads, and smoke cloud centers are established alongside a “three-step effective concealment” constraint, constructing a nonlinear model targeting effective concealment duration. Second, genetic algorithms perform encoding, crossover, mutation, and adaptive iteration to solve optimal release strategies for single-UAV-single-missile scenarios, then extend to multi-missile temporal coordination. Finally, integer programming provides initial allocations, simulated annealing searches within neighborhoods, and entropy weighting assigns multi-criteria scores. An embedded genetic algorithm evaluates strategy performance. This model rapidly approximates global optima under strong nonlinearity and coupled constraints, enhancing both effective shielding duration and resource allocation efficiency. It offers scalability, robustness, and engineering feasibility.