A Monte Carlo Analysis of Sensor Fusion Algorithm for the Estimation of Aerodynamic Angles in a Mini Aerial Vehicle

Abstract

An Extended Kalman Filter (EKF) fusion algorithm has been developed for estimating the aerodynamic angles in a Mini Aerial Vehicle (MAV). The knowledge on the aerodynamic angles would be helpful in designing control laws during the high angle of attack flights. This involves an algorithm that computes pseudo measurements of Angle of Attack (AOA) and Side Slip Angle (SSA) which are fused with typical sensor data for the estimation of aerodynamic angles. The current work focuses on testing the robustness of the estimation algorithm to uncertain aircraft parameters inherited from the manufacturing errors, assembly errors and also from prolonged usage of the aircraft. For this purpose, a Monte Carlo simulation has been carried out using number of randomly generated aircraft parameters in the flight simulation program, and the true states of the aircraft are generated. With few of these true states, a zero mean white Gaussian noise is added and the sensor data has been simulated. Using the simulated sensor data, the aerodynamic angles are estimated. It should be noted that the estimation algorithm has been provided with correct aircraft parameters, whereas the flight simulation has been carried out with randomly generated aircraft parameters. This paper brings out the robustness of the estimation algorithm to changes in the aircraft parameters propagated from manufacturing / assembly errors.