This seminar addresses the design and error assessment and analysis of a low cost terrain following system for UAVs. Such a system enables an aircraft to maintain a pre-defined vertical separation (altitude) with regard to the terrain surface over which it flies. First papers on this subject date back to the sixties of the previous century, focusing primarily on radar based terrain following systems. To address the limitations and constraints imposed by radar based systems, new approaches have emerged which use electro-optical passive sensors (cameras) and provide better compatibility to the UAV platform.
In this study a new approach is proposed to surmount the disadvantages of the previous systems and to provide a low cost terrain following solution for UAVs. This system utilizes several laser range sensors in a strap-down configuration as its measurement device. The measurements obtained from the sensors are used to generate a desired terrain following trajectory for the aircraft. A trajectory tracking algorithm is then incorporated to ensure that the aircraft remains on the desired path, thus satisfying the required altitude. The complexity and performance of the proposed system depends predominantly on the chosen layout of its sensors as well as the errors of its various components. The related estimation and control algorithms designed for the terrain following task have to address also system uncertainties and disturbances. The study discusses and thoroughly examines the various error factors affecting the system performance. The goal is to develop an analytical error model for the system thus pinpointing the different factors which may cause either improvement or degradation in terrain following performance as well as aid in the design process.
The performance of the terrain following system is evaluated using a numerical Monte-Carlo simulation environment constructed for this purpose. The simulation results clearly demonstrate that the suggested analytical error model agrees closely with the Monte-Carlo results both in trend and in values in the relevant domain. The resulting analytical error model quantifies the overall system tracking error, outlines the preferred values for various system parameters in order to achieve higher terrain following performance, and mainly suggests the optimal pointing angle for the laser designators.
