FPGA based flexible autopilot platform for unmanned systems

Abstract

Diversity in unmanned vehicle designs and control as well as diversity in existing autopilots lead to major compatibility issues among different platforms. The small number of different autopilot designs currently available in the market and their limitations, justifies the rationale for proposing design of a new modular, flexible, easy to integrate autopilot, suitable for multiple platforms. A discussion and comparison of the state of the art in autopilot research is presented, followed by the introduction of an autopilot design that will provide a novel and wellneeded processing platform for unmanned systems. The design is fully integrated with MATLAB/Simulink allowing for both simplified high level programming and hardwarein-the-loop capabilities, and provides a flexible FPGAbased platform that maximizes parallel processing and allows for analog signal conditioning that can be modified through digital communication from FPGA processor.

Execution time analysis and optimisation techniques in the model‐based development of a flight control software

“…To provide the external interfaces, two ARM Cortex‐M3 processors are used. In other projects, a field programmable gate array is often used for this purpose [17, 18]. The reason of utilising microcontrollers here was to analyse the usage of such devices for I/O tasks and to show the effort which is necessary for such an development in the context of safety critical airborne software that is regulated by DO‐178 [10].…”

An adaptively reconfigurable computing framework for intelligent Robotics

“…Osamah et al [11] propose a reliable computing framework with reconfigurable implementation of avionics and control computers based on "graceful degradation model". [12] propose a novel autopilot description using FPGA based parallel processing for signal conditioning. [6] address real time FPGA implementation of feature detection and tracking for Micro UAV systems as a solution to onboard obstacle avoidance using feature based matching.…”