Research on Projectile Range Prediction Based on XGBOOST Optimized by Improved GOOSE Algorithm

Aiming at the influence of various external factors on the range of projectile launching, and the problems of time error accumulation and precision decline in traditional range calculation. An improved GOOSE algorithm is proposed to optimize the range prediction model of XGBOOST and improve the prediction accuracy. First, the flight characteristics of the projectile are calculated by the fourth-order Runge-Kutta method, and the data set is built according to the flight information of the projectile and the prediction model is trained. The projectile range is predicted and analyzed by taking the initial flight velocity, angle of attack and wind force level as the external influencing factors. At the same time, the prediction model of XGBOOST optimized by Grey Wolf optimization algorithm, particle swarm optimization algorithm and Sparrow optimization algorithm is compared and analyzed. Verify the prediction effect of IGOOSE-XGBOOST network model. The experimental simulation shows that the IGOOSE-XGBOOST model has higher prediction accuracy than the GWO-XGBOOST, PSO-XGBOOST and SSA-XGBOOST prediction models, which is reasonable and feasible for the projectile range prediction, and can provide an important theoretical reference.