Abstract: Two algorithms are presented to solve the carrier attitude and rate estimation problem. One is based on Kalman filter and the other is based on Particle filter. In order to avoid the singular problem from Euler angle, we chose quaternion to describe the attitude angle. We carried out the data acquisition of gyroscope, accelerometer and magnetometer of mpu9250 module. By comparing the experimental data and simulation, we discussed the feasibility of quaternion Kalman filter and particle filter algorithm. The results of static and dynamic simulation show that the errors of the particle filter and the Kalman filter are similar when the carrier attitude is measured by the MEMS device, and the standard deviation of the particle filter is relatively small.