A completely autonomous helicopter may have been … having to worry about the modeling and dynamics of the helicopter. … Feron, Nonlinear Model for a Small-Size Acrobatic Helicopter … Developing Automated Helicopter Models Using Simulated Annealing and Genetic Search Instructions for Authors Developing Automated Helicopter Models Using Simulated Annealing and Genetic Search Namir Aldawoodi PhD Candidate / Computer Engineer. University of South Florida Tampa, FL 33620 Dr. Rafael Perez Professor & Graduate Program Director ..

1 INTRODUCTION The field of Autonomous Vehicle research and robotics is very active - this is due to the ability of Autonomous Vehicles to carry out Dirty, Dull and Dangerous work [1]. There are numerous applications where an automated vehicle is desirable. One such application is reconnaissance where automated pilot-less vehicles are desirable. The use of these crafts eliminates human-error due to fatigue and results in longer, more efficient missions. Another advantage lies in the fact that an automated craft is replaceable, a live pilot is not. Automated vehicles rely on controllers to carry out tasks that the vehicle is asked to complete. For example a hover controller is tasked with keeping a helicopter in the same coordinates within an acceptable error margin (5 feet for example) - so long as the helicopter is within that 5-foot sphere radius then the module is working well. Many approaches can be used to construct a module and those include PID controllers, Neural Networks, Fuzzy Logic sets among others. Each of these controllers has their own specific limitations. The method proposed would offer an alternative to modeling a controller or a helicopter (plant) when other approaches fall short or when an alternate model is needed to compare experimental results against. This method provides a mathematical formula from which analysis can be conducted or a model can be built or simulated. In other words, the method presented can apply as a primary algorithm to model a system or as a research tool to come up with alternate models to the same system. 2 BACKGROUND Compared to regular airplanes, helicopters have many drawbacks. They are costly, slow, have a low payload capacity, and poor fuel efficiency. However, a helicopter offers two main advantages: first, the ability to land anywhere and; second, the ability to fly at a low altitudes very slowly (hover). Unlike airplanes, commercial autopilot technology for helicopters is limited to level flight for the most part; taking off and landing still pose a challenge. Currently, commercial helicopter autopilots can cost upwards of $250,000. There is no inexpensive commercial autopilot software that can be used for a small kit or pilot-less helicopters so research on automated helicopters may yield new and more sophisticated autopilot modules that can rival that of fixed wing aircraft. Pilot-less helicopters are in use today; however, they are usually flown using a remote interface. The pilot would operate the craft from afar. There are two types of pilot less control, remote control and teleoperation. With remote control the plane is visible to the pilot. This is usually the case with kit planes that are flown in a limited area and are always visible to the operator. Teleoperation is where the plane is not visible to the pilot. In this case sensors on the plane (usually a camera) send information to the pilot (and crew) on the ground through an interface. This interface usually requires a team to run and also requires a large amount of bandwidth to relay data to the aircraft in a timely manner. A failure at the command center or anywhere in the communication chain can lead to loss of control and the aircraft. Remote controlled autonomous vehicles have had reasonable success, and many have autopilot as a backup system should they lose communication with ground control. Hence, a helicopter with a fully automated controller offers many advantages amongst them the ability to fly for long periods of time without human intervention,

Download Developing Automated Helicopter Models Using Simulated Annealing and Genetic Search.Pdf