基于EKF算法的太阳能无人机低成本飞控状态估计

当大展弦比太阳能无人机(UAV)采用由低成本传感器组成的飞控平台时,受限于传感器误差精度、无人机长航时、广域度的任务要求,传统数据融合算法无法实现其姿态、空速和风场长时间的准确和可靠估计.从飞控搭载的传感器测量原理出发,对测量过程的误差特性和温度影响进行建模,基于扩展卡尔曼滤波算法实现状态的可靠估计.首先,将压力传感器与惯导的数据进行融合以实现姿态估计;其次,结合无人机的布局特征将磁力计独立安装以实现航向估计;最后,融合GPS的数据进行导航估计.仿真结果表明,较传统的变增益估计算法(VGO),所提出算法的层次更分明,结果更可靠,而且可以与太阳能无人机的特征较好地结合.

State estimation of low-cost flight controller of solar-powered UAV based on EKF algorithms

When the flight controller composed of low-cost sensors is applied to a large aspect ratio solar-powered unmanned aerial vehicle(UAV), it is limited by the accuracy of the sensors, the long-endurance, and wide-range task requirements. The traditional data fusion algorithm cannot realize its accurate and reliable estimation of attitude, airspeed and wind field for a long time. Starting from the sensor measurement principle of the flight controller, the error characteristics and temperature effects of the measurement process are modelled, and a reliable state estimation is realized based on the extended Kalman filter algorithm. Firstly, the pressure sensor and inertial measurement unit(IMU) data are combined to achieve attitude estimation. Then, combined with the layout characteristics of the UAV, the magnetometer is independently installed and the heading is estimated. Finally, GPS data is merged for navigation estimation. The simulation results show that compared with the variable gain observe algorithm, the proposed algorithm is more hierarchical and the estimation results are more reliable, and it can be combined with the characteristics of the solar-powered UAV.