Research On Attitude Measurement Algorithm Of Moving Vehicles Based On MARG Sensor

AHRS is an important part of inertial navigation system,and it is the key technology of automatic driving car and intelligent transportation.Attitude measurement system based on MARG sensor,which consists of a three-axis gyroscope,a three-axis accelerometer and a three-axis magnetometer,has been widely concerned because of its low cost and small size.There are two ways to calculate the attitude angles,one is based on the integral operation of angular velocity measured by the gyroscope,and the second is based on the accelerometer and magnetometer signal.The former results will contain a lot of integral error after a long period of time.The latter results are susceptible to the linear acceleration and magnetic interference.When a vehicle is moving,the linear acceleration interference exists all the time.Therefore,how to reduce the interference of linear acceleration is the main problem in this paper.In addition,the fusion of two measurement information to achieve more accurate attitude,is also the main problem of this study.As to the magnetic interference from the vehicle,the magnetometer is recalibrated using the ellipsoid fitting method.Through the FFT(Fast Fourier Transformation)analysis of gravity projection and linear acceleration,a Butterworth low-pass filter is proposed to filter the high frequency part of linear acceleration.Positive and negative filtering method is used to achieve zero-phase digital filter.Using the odometer and gyroscope to determine the motion state of the vehicle,to estimate and eliminate the low frequency part of linear acceleration.The experimental results show that most of the linear acceleration has been eliminated after the proposed method.The processed accelerometer signal is close to the real gravity projection.Based on the error equation of inertial navigation system,an adaptive Kalman filter is proposed for initial alignment and information fusion in attitude calculating.Corresponding to different motion states,the observed noise matrix will be adjusted according to the different fusion strategy.The simulation results show that compared with the traditional method,the proposed method has advantages in the state recognition and attitude calculation.In order to verify the feasibility of the attitude measurement system,a dynamic vehicle experiment is carried out.Compared with using the original accelerometer and magnetometer signal,the attitude angle calculated by estimated gravity projection and magnetometer signal after low-pass filter,however,it still contains large calculation error.The attitude calculation of gyroscope has a high precision in a short time,but after a period of time the cumulative error will become very large.Therefore,it can not be used to attitude calculation for a long time.In this paper,the attitude measurement system overcomes the shortcomings of the two methods,and the calculation accuracy has been greatly improved.Using the MTi sensor output as accurate values,the maximum error of the three attitude angles calculated by the proposed method is 1.38°,1.07° and 1.33°,respectively.