Mechatronics systems are taking the world by storm. Manual work is being replaced by automatic machines that yield efficiency, repeatability and cost‐effectiveness. One area wherein a lot of progress has been done is unmanned ground, marine and aerial systems. Armies of the future are aiming towards ‘mechatronization’ of their fighting force which can give them a decisive advantage over their rivals in the battlefield.
Unmanned aerial systems have been successfully deployed by several countries around the world. In Pakistan, an earnest effort is being made to develop unmanned aerial vehicles (UAV) and bring them at par with the advanced UAV squadrons of the developed countries. One such initiative is being taken at the Mechatronics Department, College of E&ME, National University of Sciences and Technology. A team headed by Col. Dr. Javaid Iqbal is undertaking a project to develop an autopilot system in order to transform an RC airplane into an autonomous aerial vehicle capable of flying to designated waypoints without human intervention or guidance whilst avoiding obstacles and also have the capability to search and identify designated target buildings in an area of interest. Asad Ullah Awan and Sheir Afgen Zaheer are two research engineers working on this project.
UAV’s would be preferred over remote controlled vehicles as they are independent of human assistance. Given this, they can monitor large amounts of data for extended periods unaided by humans. Also the vehicle may be sent deep into harsh environments impossible for the human body to withstand. Furthermore, they will not be bound by the range of wireless network, which usually is restricted. Finally, the human independence factor also allows us to design vehicles of varying sizes, i.e., one may design thousands of micro UAV's, which may be used in a variety of military and commercial applications.
For this purpose, a mathematical model of the airplane was used to simulate its dynamics over a certain flight envelope. A flight control system for the autopilot was designed, tested and tuned in the MATLAB environment on the mathematical model of the airplane. Data coming from the sensors onboard the aerial vehicle is always vulnerable to noise. To tackle this problem, noisy sensory data was simulated in the MATLAB environment and a multisensory Kalman Filter based localization module was developed in order to estimate the status of the vehicle at any given instant from the onboard sensory data. The control system and localization module were integrated and tested on the dynamics model of the airplane. The modules were then implemented in C++ in the Linux operating system. A target tracking module has also been developed and tested in the MATLAB environment.
The project is currently in its implementation phase. The control system and localization modules have been ported onto a single board computer running the Linux operating system. Work has also been carried out to stream video feed from onboard cameras to a ground station terminal via a wireless link.
A research paper titled "Autonomous Control for a Kadet MKII RC airplane" has also been published in the Proceedings of the IEEE International Conference on Emerging Technologies, October 2008. The group aims to participate in the International Aerial Robotics Contest at Georgia Tech, USA in the near future.
No comments:
Post a Comment