UAV attitude estimation based on MARG and optical flow sensors using gated recurrent unit

• DOI: 10.1177/15501477211009814
• Publication Date: 2021-04-23T05:38:22Z
• Authors: Xiaoqin Liu; Xiang Li; Qi Shi; Chuanpei Xu; Yanmei Tang
• Keywords: Optical Flow; Sensor Fusion

Three-dimensional attitude estimation for unmanned aerial vehicles is usually based on the combination of magnetometer, accelerometer, and gyroscope (MARG). But MARG sensor can be easily affected by various disturbances, example, vibration, external magnetic interference, gyro drift. Optical flow has ability to extract motion information from image sequence, thus, it potential augment three-dimensional vehicles. major problem that optical caused both translational rotational movements, which are difficult distinguished each other. To solve above problems, this article uses a gated recurrent unit neural network implement data fusion sensors, so as enhance accuracy The proposed algorithm effectively make use contained in measurements also achieve multi-sensor without explicit mathematical model. Compared with commonly used extended Kalman filter estimation, shows higher flight test quad-rotor