Research On Data Fusion And Development Of Attitude And Heading Reference System For Unmanned Aerial Vehicles

The heading and attitude are the most important parameters of the flight control system for the UAV (Unmanned Air Vehicle), There are many approaches to estimate the attitude, and the most popular one for manned aircraft is using the INS. However, for a small UAV the high performance INS is too expensive and bulky due to the low-cost, weight and space constraints. With the rapid development of embedded system and the appearance of newly high-performed inertial components, it’s possible to design a low-cost and micro AHRS.The paper researches a kind of AHRS which bases on Micro Electro Mechanical System (MEMS) and Digital Singal Processor (DSP).Firstly, the paper introduces and the element theory of the the development and current situation of MEMS Strapdown Inertial Navigation theoretical base of the systematic design. Then study the characteristics of attitude estimation using different sensors and algorithms to determine for small UAV attitude estimation scheme. In order to improve the performance and relialility of the developed AHRS, two multi-sensor fusion algorithms are employed including the quaternion-based extended kalman filter(EKF) and unsent kalman filter(UKF) to fuse different sensors. Simulations results indicate that the UKF method performs well both in short-term and long-term accuracy.The whole system design of AHRS based on MEMS is given in the paper. The system consists of sensor module, ADC module, DSP and periphery port, serial communication module et al. The software system has been designed based on hardware module. Then static and dynamic experimentation are conducted to analyze the performance and the feasibility of the AHRS which has been desiganed in this paper. The preliminary testing results shows that the static measuring accuracy can be±0.1deg, the dynamic measuring accuracy can be±2deg, The validation results show that the estimated attitudes of the developed AHRS are within the acceptable region. This verifies the practicability of the proposed development procedure.In addition, the accuracy of the low-cost inertial measurement unit was influenced by vibration and the drift of the gyroscope. An attitude estimation method is presented to promote the accuracy for small unmanned aerial vehicles (UAV) powered by a piston engine. This method that fuses the gyroscopes and single antenna global positioning system (GPS) is proposed to mitigate the influence of engine vibration and gyroscope drift. Simulations results indicate that the proposed method performs well both in short-term and long-term accuracy even though the gyroscopes are affected by drift and vibration noise and the pseudo-attitude contains severe noise, and then enhance the accuracy of the attitude estimation. Laid the foundation for the next project application design.