An Attitude Estimation Algorithm for High Mobility and Magnetic Interference Environment: Experimental Results

In the traditional attitude estimation algorithm, the attitude calculation accuracy is prone to varied elements that might lead to deterioration of performances. To overcome the difficulties and to improve the gesture measurement accuracy, especially under the high maneuvering and strong magnetic interference environment, an improved gradient descent attitude algorithm based on Kalman filtering is proposed in the current paper. The error statistics in the simplified Kalman model are used to filter and compensate the sensor data, together with the gradient descent. Then the data fusion and attitude solution are carried out. The results of the maneuver experimental test showed that the solution accuracy in the above environment is significantly improved compared with the conventional complementary filter algorithm. The algorithm seems to be also simple to implement, the computation burden is also low, and thus has potential in varied practical applications.