(Center for Integrated Spintronic Devices, Hangzhou Dianzi University, Hangzhou 310018, China)

 Abstract: Aiming that the present most of the flight control system of the attitude information is incomplete and the data fusion algorithm is complex, the posture data acquisition, sensor data features, data fusion and flight control models of quadcopter were studied, multi sensor data fusion characteristics were summarized, detailed description of the implementation process of the complementary filter algorithm was made, a quadcopter control system was designed using complementary filter fusion algorithm. In this control system, gyroscopes, accelerometers and electronic compass were used as a sensor group to detect the attitude information, control algorithm was achieved by the classic PID controller and the flight control system was completed on the STM32 platform. Finally, in the production of finished hardware, the effects of data fusion algorithm were compared and analyzed. The results indicate that the date of MEMS sensor and GMR sensors can be fused better, can get high reliability attitude angle data, and also can well satisfy the requirements of the four axis aircraft flight control.

 Key words: quadcopter; quaternion; GMR; complementary filter; micro electro mechanical system(MEMS)