I have also been writing statements for a few interesting projects for simulated drones, that could be taken up as FYPs at the undergraduate level. For a more detailed review of these projects, please review them on the following links:

• Software Dovetailing of ECJ With UAV Playground.
• UAV Playground in Matlab.
• Dovetailing of ECJ With UAVPlayground Using Matlab or Octave.

These are all nice projects. An aspect worth noticing about all of these is that they involve UAV Playground. UAV Playground is a UAV simulator that runs in conjunction with the FlightGear simulator and Google Earth. The reason I chose it was that it is written in Java, and I wanted to port it to Matlab somehow. These projects are nice and are worth pursuing.

However, here I want to introduce a very project that can be accomplished solely with Matlab, with minimal involvement, whatsoever, with another, or any other, programming language, whatsoever. The reason for this digression for this has a basis in my recent discovery of Flight Simulator Interface for Matlab. With this tool, you can run FlightGear simulator with Simulink models. Have I introduced FlightGear to you in this tutorial? I must have done that elsewhere, of course! Stated shortly, FlightGear is an open source flight simulator, that is possibly also the best in its niche domain. With this interface, you can run FlightGear models through Matlab and/or Simulink. This should lead to a lot of time-saving for you if you want to work with flight simulators or development of UAVs. So following are the things that I expect you to do in this tutorial.

1. Download and install FlightGear on your machine, by following the guidelines on this Matlab page.
2. Run FlightGear with Simulink models and see how you can gain experience with using FlightGear through Matlab.
3. Browse and look for the machine learning models that you have available for Matlab.
4. Try to figure out a way that how you can integrate these, or similar, algorithms with the flight simulation interface.

You will find the following articles particularly useful for this project.

A Survey of Free Software for the Design, Analysis, Modelling, and Simulation of an Unmanned Aerial Vehicle

The objective of this paper is to analyze free software for the design, analysis, modelling, and simulation of an unmanned aerial vehicle (UAV). Free software is the best choice when the reduction of

Automatic control education using Flight Gear and MATLAB based virtual lab

In this paper, a virtual lab based on MATLAB and FlightGear flight simulator is discussed and developed. The virtual lab consists of number of experiments that enhances the understanding of fundamental concepts of classical control for undergraduate students.

An Error Occurred Setting Your User Cookie

In general, only the information that you provide, or the choices you make while visiting a web site, can be stored in a cookie. For example, the site cannot determine your email name unless you choose to type it.

NASA HL-20 with Optional FlightGear Interface – MATLAB & Simulink Example

This model shows how to model NASA’s HL-20 lifting body modeled with Simulink®, Stateflow® and Aerospace Blockset software.

FlightGear Route Manager – File Exchange – MATLAB Central

File exchange, MATLAB Answers, newsgroup access, Links, and Blogs for the MATLAB & Simulink user community

UAV Simulation Environment for Autonomous Flight Control Algorithms – ProQuest

This thesis presents the development of a UAV simulation environment for the design, analysis, and comparison of autonomous flight control laws. The simulation environment was developed in MATLAB/Simulink, with custom map generation software and FlightGear 3-D visualization. Graphical user interface of the simulation environment is user-friendly and all available options are discussed in detail.

Co-Simulation of Matlab and FlightGear for Identification and Control of Aircraft

The paper outlines the development of a co-simulation solution of Matlab and Flight- Gear in which the communication between these programs is done via UDP without needing further toolsets. The simulation and rendering is done by FlightGear. Flight measurement signals are sent to Matlab, which in turn sends back actuator input values computed by a flight control system.

Howto:Design an autopilot – FlightGear wiki

Why not make things slightly easier, before we start working on the autopilot? Since we will be updating the autopilot hundreds of times, making tiny adjustments each time, we don’t want to reload FlightGear every time. Therefore, we assign an “reload autopilot configuration” action to a key on our keyboard.

Flightgear Integrations

Audio: – social audio

The following video shows how FlightGear works on a user’s desktop.

FlightGear: Seneca II Flight

FlightGear: Seneca II Flight

You can find a lot more FlightGear resources on the following page.

More Flightgear Resources

Photo by jinterwas

ECJ is a well known evolutionary computing software that is written by Sean Luke at George Mason University. It is written entirely in Java. There is another sister project of ECJ that is MASON. MASON is for modeling mobile agents in an artificial environment. MASON dovetails very well with ECJ. This means that ECJ can be used to evolve agents for MASON which have reasonable intelligence in a meaningful sense to maneuver themselves in the MASON environment.

The idea is to closely inspect how MASON is dovetailed with ECJ. Once this is understood, the next thing to be done is to figured out that how some other software can be dovetailed with ECJ. In particular, it would be really nice to figure out a way to dovetail TORCS with ECJ. Similarly, it would be nice to figure out a dovetailing of UAV Playground with ECJ. This can be a nice FYP and is expected to have many benefits.

What is dovetailing by the way? The term merits some explanation. Dovetailing, as the name suggests, is to integrate two software together so that one runs the other.

Photo by fusion-of-horizons

A good training software could be the ECJ package. ECJ is an evolutionary computing software package written in Java. It has quite a few nice algorithms implemented in it. A few examples are evolution strategies, genetic programming, genetic algorithms and particle swarm optimization.

A good gaming engine could be the TORCS simulator. TORCS stands for the open racing car(s) simulator and is used by the famous “The Simulated Car Racing Competition”. Other gaming engines could be thought about in which you can make objects and they act in a smart way in some meaningful sense. Software simulators for physical objects and devices could be counted in the same category.

One nice software simulator is the UAV Playground. With the UAV Playground you can create various types of drone plane as software objects in the simulator and have them fly (in the same simulator of course).

Flying an object with a joystick in a simulator is one thing. Another important thing is to the objects (drones in this case) by themselves in an intelligent manner. This is the challenge.

As discussed above a training software like ECJ can be used to accomplish this goal. The training software and the simulation software have to be integrated in order to accomplish this, however.

The challenge is to write an application integration software that:

1. Integrates ECJ and UAV Playground.
2. It should be easy to create members in ECJ that correspond with the controllers of the drones UAV Playground. This is to say that the members of ECJ should have a representation that is amenable with the representation of the controllers of the drones.
3. The integration software should preferably be written in Java and/or Matlab.

Photo by marfis75