The Design Of Heading And Attitude Measurement System Based On Multiple MEMS Sensors

The heading and attitude measurement system based on multiple MEMS sensors is a kind of micro measurement system, which is used to export body’s heading and attitude information in a variety of model aircraft flight control platforms, wearable devices...etc. The heading and attitude measuring precision of this system is an important index to evaluate its quality, and the key technology is the multi sensor data fusion algorithm. The commonly used methods are kalman filter, gradient descent method, complementary filtering method...etc, but these algorithms can lead to deterioration of the accuracy of attitude in complex environment.According to the heading and attitude measurement system in complex environment, this paper implements an improved algorithm of extended kalman filter based on UDUT factorization and adaptive noise evaluation. This algorithm increased stability calculation through UDU factorization, and adjust the variance matrix of sensors dynamically through real-time detection on the data characteristic of sensors, and that increase the accuracy of the system.In the verification experiment, compared in a complex environment by using kalman filter and optimized kalman filter, the algorithm can indeed improve the accuracy of heading and attitude measurement system under complex environment.Finally, this paper present a heading and attitude measurement system based on three axes MEMS gyroscope sensor, three axes MEMS accelerometer sensor and MEMS three axes magnetometer sensor, which estimates the heading and attitude angles through the optimized kalman filter. This system’s static error is less than 0.3°, and dynamic error is also less than 2°, it has the ability to work in a complex environment. So this system can be used as the embedded system platform to development user application.