Performance Test 3

The first technique implemented was accelerating the AEV using the celerate() command, then rotating the prop and running a pull brake that slowed the AEV to a stop. Initially the team was happy with  the results of doing this, but then as Team P began to focus more on efficiency other code variations were considered. Next, the team decided to implement motorSpeed() rather than celerate() after a stand still, because the track in room 224 has a slumped spot in the beginning that the AEV had trouble negotiating when accelerating from a low power level. Also, after testing the two code variations against each other the code that had the motorSpeed() command used less overall energy than the one with celerate(). Because of this Team P decided to do away with the celerate() command and focus on motorSpeed().

On the second day of lab, the team then continued on improving the code design by examining whether or not a rotation of the servo and then a pull brake was actually necessary when slowing into the cargo and the finish. Coasting lightly into the cargo was found to be a better option to reduce energy usage since it is less precise of a stop than properly triggering the gate sensor. The team achieved this by simply reducing the input power of the AEV when it was going from the gate to the cargo, and from the gate to the start. To show the increased efficiency of coasting the AEV vs braking it, Team P did a run where the AEV coasted into the cargo, and then was manually braked at the end of the run. It was seen that between 30 and 33 seconds not a large amount of energy was used (coast); only a small amount was used to rotate the servo. Between 63 and 65 seconds a large spike of energy usage can be seen which was where the AEV was braked. This was only done as an example to show the difference in energy usage. Team P then changed the code so that the AEV would coast in both portions of the code; however, Team P ran out of time before data of a run with both coasts could be downloaded. 

The example data used a total of 178 Joules of energy with an AEV that weighs 280 grams. Team P was happy with this result. The J/kg ratio was calculated to be 635. Maggie seemed very surprised at how low the ratio was which was a good sign. That was also with a manual break at the end of the run, so the Team knows the ratio will be even lower on final test day.

On the third day of testing, the team attempted to adapt the code to fit the track in room 308. Team P did so by running multiple attempts and adjusting the code slightly each time. One adjustment made was when the servo rotated while approaching the gate. The team observed that when the rotation occurred along the curve in the track the AEV lost significant momentum and didn’t reach the gate. The timing was delayed to account for this. Other adjustments made included changing the motorSpeed() arguments and the marks that were used to trigger rotations. This process took up the whole lab period, and Team P ended up not completing a full run once. Because of this, Team P was again unable to download data of a coast into the cargo and into the finish line.

Graph 1: Run Data

Figure 1: Final Design