SELF-PROGRAMMED MOUSE

Self-Programmed Mouse

Design Concept and Background:

There are three goals that I am aiming for when I was about to make this controller. I was planning to use the servo as the accelerometer however, due to parts failure during testing, I decided to replace it with a potentiometer to control its speed. By using the potentiometer as the accelerometer, it is actually a little bit easier for me to control it since both potentiometers are placed pretty close. Aside from that, using a potentiometer allows me to design a simpler controller which is the second goal of mine. Lastly, my third goal is to make the most accurate steering mechanism possible.

In my final design, I ended up using 2 potentiometers. One for the steering mechanism and the other for the accelerometer. I decided to place the potentiometer a little bit more isolated so that I can control those two easily. I decided to place the steering potentiometer on the left to mimic the layout of a regular gaming controller which allows the user of the mouse to have a natural feel when controlling it.

Materials Used:

• Breadboard

• Potentiometers

• Arduino Uno

• Wires

Softwares Used:

• Arduino

• Matlab

Testing Development

*Control law: Closed-loop (The user will be the one who gives the controller the feedback)

Performance Criterion:

In this project, I tried to optimize the throttle which controls how long will it take the robot to reach one corner from the middle. The main purpose of this optimization is to create a more viable and responsive robot. The goal is to reach one corner from the middle within 25 to 30 seconds.

Experimental Parameter:

In the MATLAB commands, I specifically only play with the impulse response for the experiment and leave KP_values and firing piston rate at the same value. Some experimental values that I used for the impulse response are 8500, 10500, 12500, 14500, and 15500. To create a more accurate observation, 10 trials are done for each data set.

Experimental Results:

Based on the result from the graph, as the value for impulse response increases, it will take a shorter amount of time for the robot to reach the corner. However, instead of using the highest testing value, I decided to use 12500 because it is the most stable one even though it is not the fastest. From the standard deviation itself, it can be seen that the 12500 is in the middle, not too big yet not too small. I decided not to go with 8500 because even though the standard deviation is small, the time for it to reach the corner is the longest hence, I think using 12500 as the impulse response value will produce the best result.