An Optimization Model and Computational Framework for UAV Smoke Deployment Using Simulated Annealing and Genetic Algorithms

Abstract

This paper addresses the issue of smoke screen formation for drone-deployed missiles by proposing a computational method integrating geometric modeling and intelligent optimization algorithms. Kinematic models for the drone, missile, smoke grenade, and smoke cloud are established within a three-dimensional coordinate system. By constructing a geometric occlusion criterion between the missile's line of sight and the spherical smoke cloud, the shielding process is transformed into a spatial intersection problem between a line and a sphere, enabling the calculation of effective shielding duration. Building upon this foundation, a nonlinear optimization model is established with maximum effective concealment duration as the objective function. This model incorporates variables such as flight direction, flight velocity, smoke grenade release time, and detonation time. For single-UAV deployment, parameter optimization is achieved using the simulated annealing algorithm. For multi-UAV coordinated deployment, a genetic algorithm is employed to ensure seamless coordination between multiple smoke screen coverage intervals. Simulation results demonstrate that this method effectively extends concealment duration while exhibiting excellent stability and adaptability.