Data Fusion of Gyroscope and Magnetometer to Estimate the Attitude of High-Speed Projectiles Based on Particle Filter and PSO

In this paper, the attitude of high-speed projectiles has been estimated usingdata fusion of magnetometer and Micro Electro Mechanical Systems (MEMS) gyroscope. MEMS gyroscopes have the high error for high speed. Also, magnetometers have low accuracy due to the presence of Non-Earth magnetic fields. For this reason, data fusion of magnetometer and MEMS gyroscope have been suggested. Due to the nonlinearity of the system equations and observation, a nonlinear estimator must be used. The developed Kalman filter inserts an error by ignoring the high order sentences of Taylor's expansion, which cannot be ignored in fast nonlinear systems. Unlike the Kalman filter, the particle filter has good results for nonlinear systems. The biggest weakness of this filter is its high computational time, which limits its applicability. To reduce the computational time of particle filter, a particle swarm optimization algorithm has been used. The simulation results were evaluated using 100 samples of the test, which illustrates the desirable performance of the combined particle filter with the particle swarm optimization algorithm in the data fusion of gyroscope and magnetometer information in the estimation of angles.