Home

Welcome to the wiki page of Path-Tracking-Simulation

Checklist

Here is the list of the ongoing tasks and the completed tasks

PID Controllers:

• Normal PID
• MIT-Rule
• Adaptive PID
• [PID for path tracking]
• [Adaptive PID for path tracking]

Geometric Models for path tracking:

• Pure pursuit
• Stanley model

Note- The stanley model is working correctly on circular paths but it is not working correctly on sinusoidal paths.

Optimal Controllers:

• LQR for steering with constant speed

Others:

• Codes for path generation which will be used for bench-marking

1. Straight line with the bot at some angle from the path
2. lane Switching (Bot will be placed on the path)
3. Circular Path
4. Sinusoidal Path (or Cubic spline)

Meetings

21-08-2018

• Introduction to Github
• Properly formatting the python codes using PEP8 convention
• Properly commenting the codes with docstrings, and mentioning the author, references, convention and heuristics in each code
• Pushing all the codes immediately after they are properly formatted and commented
• Maintaining a WIKI page which includes the checklist(timeline) and Daily updates. Also add the subpages on Authors, Convention, References
• Detailed explanation of Docker
• Need to shift to CPP from python later on and also start learning docker so that it can be used later on.

29-08-2018

• General work on the current work
• Test PID, Adaptive PID on car and then Pure Pursuit
• Fix the error in Stanley code
• Start the course "control for mobile robots" and "control-bootcamp"
• Reference books- a) Ogata b) Nice c) B.Kvo
• LQR model
• config files for different paths which will be used in benchmarking
• Explanation of fixed gain controllers (pole-placement method)

5-9-2018

• Follow a naming convention for the scripts
• Refer to Hitchhiker's Guide To Python
• Make the topic names meaningful
• Add an option for inputting the path offset
• Make a running instructions read-me file
• Minimize the use of Global variables
• Remove the print statements and use roslog instead

4-10-2018

• Port path-tracking code to C++ from Python
• Conduct unit tests on standard paths like straight line, circular, sinusoidal, eight (8-figure) and record their data
• Using data from unit tests, plot cross-track error, centripetal acceleration and jerkiness (derivative of acceleration)
• Make use of bag files (recorded earlier during testing) to conduct unit tests
• Make use of state space in the code
• Complete the course "Control of Mobile Robots"