Optimization Method for Launch Vehicle Attitude Control Parameters Based on Bayesian Optimization

The paper presents a Bayesian Optimization-based method for designing attitude control parameters of Launch Vehicles. It proposes an algorithm for optimizing control parameters using small sample simulation data to construct a Gaussian process regression model. This approach significantly reduces simulation time and costs compared to traditional methods. The paper demonstrates the effectiveness of the method through experiments, showing its ability to quickly find appropriate control parameters, meet margin design criteria, and offer significant advantages in simulation costs over Particle Swarm Optimization algorithms.