Background and Purpose:
Our team is responsible for writing a section of the Concept Design Report required for the Autodrive competition. We are trying to implement new methods for our car to navigate without GPS. Another use was getting new members up to speed for understanding where our team stands in this competition, and how our car navigates itself.
Our goals for the Fall semester 2019 were to improve the robots navigation method in some way. That requires research into methods that can be used, what devices we have access to, and ways that they could possibly be implemented. We had to figure out ways to test our methods without access to the vehicle to hammer out any kinks before we could try to implement it on the car. We are currently trying to implement Dead Reckoning on the car for it to navigate without GPS for a limited period of time.
- Presentation on Localization Concepts.
- Simulation of Dead Reckoning in MATLab.
- Training for the vehicle.
This shows where we are currently at, as of 11/22/2019, with our simulation of dead reckoning. It shows the position over time estimate of the car in the top right graph.
The top left graph shows the x position and the y position graphed separately with time on the x axis. The middle graph shows the derivative for both x and y graphs from the top left. This data is converted from wheel tick data we received from the sensors on the car.