Research On Magnetic Interference Compensation And Information Fusion Technology Based On Magnetometer/MIMU Under Complex Environment

With the development of navigation technology,micro electro-mechanical system has many advantages,because of their low accuracy and the accumulation of errors over time,these defects limit its application.For the above reasons,the information of single navigation is not reliable.With the development of the geomagnetic navigation technology,the geomagnetic vector is unique and the magnetic field can only be measured by the magnetometer,so the influence of the magnetometer is only the magnetic field and not affected by other interference,and the error will not be accumulated over time.At the same time,the equipment can not calculate the complete navigation information independently.In this paper,we will remove the environmental error and the error itself which influence the magnetometer,these two system are combined to form an integrated navigation system,so as to solve the navigation information more accurately and effectively restrain the divergence of the error.In the study of the subject,we do not use GPS,which is due to GPS interference by the environment in many places and the signal we accepted is not accurate,such as: forest,indoor and so on,especially in wartime GPS-free signal,single inertial navigation will appear large deviation.Therefore,this subject uses the magnetometer and MIMU system composed of MEMS sensor to fuse the information,which is used in the ideal civil navigation field to improve the system precision and increase the precision endurance.In this paper,the magnetometer is calibrated to remove the interference by itself.In the complex environment,the calibration is carried out,so the new error model is established.Calibration experiment is to use the three axis turntable to select the appropriate location for data acquisition,and the choice of the location need us analyze and compare to determine the final program.In order to simplify the algorithm and eliminate the interference of other uncertain factors,we need around an axis to rotate 180 degrees and then adding it.After the calibration of the sensor in the equipment and before it used,we need to eliminate the interference of the environment,so it is necessary to carry out magnetic interference compensation.In the course of the study,the paper analyzes and compares the advantages and disadvantages of the best ellipse fitting algorithm,and the principle of optimal ellipse fitting under constrained conditions is deduced.The algorithm of optimal ellipsoidal fitting based on the constraint condition is proposed and further improved to make up its disadvantage,until the optimal interference matrix parameters can be solved.The compensated magnetometer output is the magnetic field value of the carrier system in the ideal state,and with the output of MEMS sensor information to fusion,which can improve the navigation precision and restrains the divergence degree of the system.As the sensor data error is larger,linearization directly can not be adopted,so the EKF is used for filtering.In the study,we analyze and compare the pure inertial navigation filter model,and improve the algorithm which has defects.In the integrated navigation system,the magnetometer information is joined and the error quaternion is added as the state vector to reconstruct the model equation.The information fusion can effectively improve the navigation precision and restrain the divergence.This paper uses the laboratory turntable and MEMS for data acquisition,and choose different locations for the calibrate experiment.The magnetometer is rotated sufficiently to obtain an ellipsoid model to fit the solution parameters for magnetic interference compensation.Finally,the design experiment will be magnetometer and MEMS information fusion solution for navigation information.The test results show that the integrated navigation system can validly enhance the system precision and restrain the divergence.