NXTway-GS (Self-Balancing Two-Wheeled Robot) Controller Design

it's Excellent Simulation to NXT robot and Documentation is perfect.

> Anand
This file is archived with "Compressed (zipped) folder" shortcut of Windows XP. Please try to decompress it using WinZip or other utilities.

Hello,
 
I'm trying to enhance (expand) the autonomous mode that controls the NXTway, this by implementing the pure pursuit algorithm, which is executed when these option is chosen.

During the simulation, I found to have some doubts:

1. Why to calculate the current position of the robot you do not use the equation 3.1 on page 7 of document (revision 1.1)?, but instead you use this one:
 
Z = integral (thetadot * R * cos (phi)) + initial Z;
Y = integral (thetadot * R * sin (phi)) + initial Y

I know that you renamed the coordinates of the robot (pg 56), but despite of the change of coordinates I think the equations are not equivalent.

2. As I could see in the file param_controller.m the maximum linear velocity is 0.3 [m / sec], but you do not put the equivalence with the angular velocity, so: If I put in the gamepad phidot_ref=10, how much it is equivalent in angular velocity?, what I see is that what you do is to add and subtract power to the motors (with a maximum of 25 pwm).

I really appreciate your attention. Thanks in advance, any help you can give me and I am attentive to your answer.

Hello,

has somebody implented an observer based strategy for the non measured states? Great work by the way, but from the view of modern control theory an observer based method would be even better.

Philipp, i have completed the controller becoming both dynamics into one, and close the loop control in the wheel dynamics, but i calculated the remain states from the on-board sensors. the controller can track lineal and angular velocities. but the idea of an observer was made by Takachi in his early nxtWAY in with use a ligth sensor to determinate the states (ecrobot exmples).