Autonomous Navigation System Using a Fuzzy Adaptive Nonlinear H∞ Filter

Sensors

Fariz Outamazirt¹  Fu Li¹  Lin Yan¹ 

[1] School of Automation Science and Electrical Engineering, Beihang University, 37 Xueyuan Road, Haidian District, 100191 Beijing, China; E-Mails:

关键词: UAV localization;    sensor data fusion;    Extended Kalman Filter (EKF);    Nonlinear H∞ (NH∞);    Fuzzy Adaptive Nonlinear H∞ (FANH∞) filter;   

DOI  :  10.3390/s140917600

来源: mdpi

【 摘 要 】

Although nonlinear H∞ (NH∞) filters offer good performance without requiring assumptions concerning the characteristics of process and/or measurement noises, they still require additional tuning parameters that remain fixed and that need to be determined through trial and error. To address issues associated with NH∞ filters, a new SINS/GPS sensor fusion scheme known as the Fuzzy Adaptive Nonlinear H∞ (FANH∞) filter is proposed for the Unmanned Aerial Vehicle (UAV) localization problem. Based on a real-time Fuzzy Inference System (FIS), the FANH∞ filter continually adjusts the higher order of the Taylor development thorough adaptive bounds (δ_i) and adaptive disturbance attenuation (γ), which significantly increases the UAV localization performance. The results obtained using the FANH∞ navigation filter are compared to the NH∞ navigation filter results and are validated using a 3D UAV flight scenario. The comparison proves the efficiency and robustness of the UAV localization process using the FANH∞ filter.

【 授权许可】

CC BY   
© 2014 by the authors; licensee MDPI, Basel, Switzerland.