Wind Field Estimation and Its Utilization in Trajectory Prediction

Abstract

This work develops a method for onboard estimation of wind field with spatial and temporal variation based on local wind vector estimation and/or measurements from multiple aircraft flying in the same airspace. Aircraft flying in the same airspace of operation are considered airborne wind sensors scattered over the airspace because of the fact that aircraft carry along with them wind information inherent in their dynamics and kinematics. The onboard wind field estimation is formulated in the framework of parameter estimation based on various wind field models, which are different function of position and time. The online wind field estimation is utilized in trajectory prediction of aircraft flying in spatially and/or temporally varying wind. Various simulation cases are presented to demonstrate the feasibility of wind field estimation and the benefit of using such information in trajectory prediction.

Real-Time Autonomous Obstacle Avoidance For Unmanned Aerial Vehicles

“…For the considered trajectory generation, only the motion of the vehicle is of interest. Therefore, as suggested by Feldman ([87], Chapter 2), the dynamic model with assumptions that a fixed-wing UAV is a point mass model, in symmetric flight, with no any lateral force and no sideslip, and without regard to wind field from [88,89,90] is adopted as follows, where M is the mass of the aircraft, L (lif t f orce)…”

“…The wing surface area is S, A is the aircraft's aspect ratio given by A = b 2 S , where b is the wing span, e is the Oswald e ciency, and ⇢ is air density. In Equation (3.1), the e↵ect of thrust (T ) has been added to the model derived in ( [89], Chapter 3). In (3.1) the thrust line is assumed to be in line with the zero-lift line (↵ = 0) of the aircraft.…”

“…For our case of comparison between a dynamic and a kinematic model for online trajectory generation, the point mass model of an aircraft is su cient, since the interest of this comparison lies in understanding the motion (position) of a vehicle for obstacle avoidance at a high level. Therefore, by adding the e↵ect of thrust, a dynamic model of an aircraft derived in references [88,89] and [90] in the presence of a wind field assuming a point mass model, is written as ⇡Ae . The wing surface area is S, A is aircrafts aspect ratio given by A = b 2 S , where b is the wing span, e is an Oswald e ciency, and ⇢ is air density.…”

Air Data Estimation by Fusing Navigation System and Flight Control System

“…Usually, this is provided by the navigation system. Thus, new thoughts, including aiding the navigation system with an aerodynamic model (Lyv et al, 2015; Rhudy et al, 2013), estimating air data parameters using navigation information (Kampoon et al, 2015; Kargaard et al, 2015), and taking stock of an aerodynamic model based on the navigation parameters and air data (Chowdhary and Jategaonkar, 2010), have attracted considerable attention. As the ground speed (speed of the aircraft relative to geographic coordinates) is the vector sum of wind velocity and true air velocity, some researchers have estimated air data parameters after the wind speed estimation (Nebula et al, 2013; Cho et al, 2011).…”