四旋翼飞行器姿态稳定性优化研究

The Study of Stability Optimization Depend on Four_Rotor Aircraft

姿态稳定在整个四旋翼飞行器系统中起到了关键的作用,但由于模型具有高度非线性,捷联惯导测量单元中陀螺仪积分误差大,加速度在时变环境误差大等问题。为了得到更加精准的姿态角,在扩展卡尔曼滤波(EKF)的基础上研究姿态性能,要解决扩展卡尔曼滤波算法中的预测误差P的发散和过程噪声矩阵Q和测量噪声矩阵R的设置的问题,提出了UD分解滤波结合自适应过程噪声矩阵Q和测量噪声矩阵R调整。经过仿真和实际测试与互补滤波和常规EKF算法对比,提出的方法在估计精度和收敛速度都明显好于互补滤波算法和卡尔曼滤波算法,能够在时变的环境下得到更加精准稳定的姿态角。

Attitude stability plays a key role in the whole four-rotor aircraft system, while the model has some questions, such as highly nonlinear, large integral error of the gyroscope in the inertial measurement unit, and the ac- celeration error in time-varying environment. In order to get a more accurate attitude angle, the attitude optimization algorithm based on the extended Kalman filter （EKF） is studied. To resolve the prediction error divergence of EKF and the settings of process noise matrix Q and measurement noise matrix R, the adjustment UD decomposition filter combining with the adaptive process noise matrix Q and measurement noise matrix R is proposed. Compared the result of simulation and actual test with complementary filter and conventional EKF algorithm, the proposed method is signif- icantly better than the complementary filter algorithm and Kalman filter algorithm in the estimation accuracy and con- vergence speed, which can get a more accurate and stable attitude angle in the time-varying environment.