Finally …

Last Thursday, I already submitted my final report.

The D-day is tomorrow, the day to defend my report, September 18th 2008 on 8 am.

Wish me luck …

News: Autonomous Helicopter from Stanford

Professor Andrew Ng, center, and his graduate students Pieter Abbeel, left, and Adam Coates

Stanford computer scientists have developed an artificial intelligence system that enables robotic helicopters to teach themselves to fly difficult stunts by watching other helicopters perform the same maneuvers.

The result is an autonomous helicopter than can perform a complete airshow of complex tricks on its own.

The stunts are “by far the most difficult aerobatic maneuvers flown by any computer controlled helicopter,” said Andrew Ng, the professor directing the research of graduate students Pieter Abbeel, Adam Coates, Timothy Hunter and Morgan Quigley.

The dazzling airshow is an important demonstration of “apprenticeship learning,” in which robots learn by observing an expert, rather than by having software engineers peck away at their keyboards in an attempt to write instructions from scratch.

Stanford’s artificial intelligence system learned how to fly by “watching” the four-foot-long helicopters flown by expert radio control pilot Garett Oku. “Garett can pick up any helicopter, even ones he’s never seen, and go fly amazing aerobatics. So the question for us is always, why can’t computers do things like this?” Coates said.

A Little Bit About X-Plane

X-Plane is a proprietary flight simulator for personal computers produced by Laminar Research. It runs on Linux, Mac, and Windows. X-Plane is packaged with other software to build and customize aircraft and scenery. The packages are X-Plane (the actual flight simulator), Airfoil-Maker (to make airfoils for your aircraft if you would like to make your own planes), Plane-Maker (to make your own planes), World-Maker (to make your own scenery to fly in if you like), and Weather-Briefer (to get a weather-briefing before your flight if desired). X-Plane also has a plugin architecture that allows users to create their own modules, extending the functionality of the software.

Problem with Serial Interfacing to Simulink

I’m trying to do a HIL(hardware-in-the-loop) simulation with ATmega8535, as a digital controller and a virtual plant that implemented in Matlab Simulink. The AVR and Simulink are communicating using serial port. I used the Instrument Control Toolbox in Matlab to send and receive binary data via serial port to the AVR. I also include the direct simulation, without using the Instrument Control Toolbox, to compare both the results.

Finish It!

Today, I found an inspiring article in here. The article told us that it’s good not to hesitate in starting to do something, just like NIKE’s tagline, “Just do it”. But on the other side, it is also important to finish the things that we have already started. According to the writer, most of our project have a “S-curve” characteristic.

The D-day was Passed

The D-day was passed. It has been almost 2 years since this research started. And it’s not finished yet.

Aiming for June …

Yesterday Meeting @ CENTRUM

New insight:

• the flight data : don’t detrend (cruise data —detrend—> hover data ) it, use the proper data for each flight condition, use data in its trim condition
• coupled effect of the model structure
• the model’s initial value list given

System Identification

In this project, we used the parameter identification approach instead of using the theoretical model. So we use the System Identification Toolbox 7.0 in Matlab R2007a as a tool for that kind of approach. The first step is to prepare the flight data for identification. In this step, I removed the data means by using ‘detrend’. Then, the detrended data was divided into two parts. The first is for the estimation data and the second one is for the validation data.

30 days to go …

February 7th 2008 is the D-day.

Selecting a Model Structure

This week I studied some model structure for system identification process. And I found a very helpful source in internet to this subject (thanks to Mr. Google). The tutorial is from National Instruments. These are some citation of the tutorial.

The ARX model is the simplest model incorporating the stimulus signal. The estimation of the ARX model is the most efficient of the polynomial estimation methods because it is the result of solving linear regression equations in analytic form. Moreover, the solution is unique. In other words, the solution always satisfies the global minimum of the loss function. The ARX model therefore is preferable, especially when the model order is high.