Fitting method of spinning projectile tri-orthogonal geomagnetic output based on improved particle swarm optimization algorithm

Obtaining the characteristics at a characteristic point of the outputs is a key step to the geomagnetic attitude measurement method of the spinning projectile. However, actual outputs usually have some noise that causes the characteristic points to deviate from the theoretical position or produce multiple fake characteristic points, resulting in the increase of solution error and even the failure of solution. In addition, the coning motion and inaccurate initial alignment of the spinning projectile increase the number of unknown parameters and the computational complexity. In this study, several improved particle swarm optimization strategies are proposed. The actual outputs are fitted to the geomagnetic output equations under the coning motion, and the supervised learning effect of each strategy is analysed and compared. The algorithm can be flexibly adjusted according to different needs in actual use by selecting appropriate strategies, which has a wide applicability in data fitting.