Performance test 3 being the final project, we would have to do all of performance test 1 and 2, then move the caboose up to the gate again, wait, then go through and end up at the starting dock, perfectly in the starting dock area. We used our code for performance test 1 and 2, and then from their had to manipulate our data to thus take the caboose back up. We realized there was a lot of power that would be required to move the caboose, so we wanted to utilize as much coasting as possible. We soon found out that it was nearly impossible to not have power braking involved. The coasting with the caboose was extremely inconsistent, so we added to our code some power braking. This choice brought up our cost, but also raised our accuracy.
When performing this test, we had two successes and one fail. A score of a 38, 40 and a score of a 24. There was no difference in code between these tests. They were all the same exact code, just different results because of the pure inconsistency of the AEV. For some reasons it would go further than normal, then not go nearly as far. It was a frustrating problem because we would think we got it right, then a drastic change would occur.
We found that there was no way to eliminate these inconsistencies so we decided to hope for the best while performing the tests. Our first attempt succeed up until it didn’t go far enough to get into the gate while attached to the caboose. We then had a complete success with our second attempt. Then our last attempt was a complete fail as the AEV while attached to the caboose blew through the second sensor in the gate.
Our data came up satisfying though. We used a total of around 200 J of energy and a average of 48 seconds to complete. This resulted in a total budget of $570,000 and a average of 39/40 on accuracy. We believe that with more time we could get the accuracy up, and the power and time down.