[New Feature] Model Predictive Control(MPC) for Path Tracking

[ShisatoYano]

Description
With the recent merge of MPPI(#48), I'd like to propose the implementation of a Model Predictive Control (MPC) based path tracking algorithm to provide a comprehensive comparison of tracking performance.

This allow users to compare the strengths and weakness of deterministic optimization (MPC) vs stochastic sampling (MPPI).

Tasks

• Optimization Problem: Formulate the cost function (tracking error, control effort, smoothness) and constraints.

• Solver Integration: Install CasADi and do-mpc in docker container and are used as the solver.

• Sample Simulation: Create a script to follow a reference course.

Reference code
When you implement the code, you can refer to the existing code of MPPI (#48).

[namwoam]

Hello, I would like to take on this. Could you assign it to me?

[ShisatoYano]

Great! I’m excited to have you on board for this feature.
Looking forward to your PR! Let me know if you have any questions during the implementation.

[kunj1302]

Hi Shisato , I would like to have a go at this. I have coded a MATLAB version of this problem using Casadi and iopt, so I think I can work on this.

[ShisatoYano]

@kunj1302 Thank you for your proposal! This issue has already been assigned to @namwoam, so if you have any ideas to implement any other algorithms, you can create a new issue for it.

[kunj1302]

Sure @ShisatoYano . I will spend some time with the code base and think about it.

[mohitk3000]

Hey, is someone working on this ? if this issue is not assigned I can take this.

[ShisatoYano]

Hi @mohitk3000, some contributors suggested working on this issue but their PR has not been sent to me yet. If you have time and can take this issue, I wanna assign you!

[mohitk3000]

@ShisatoYano here is my PR .
Looking forward to your comments.
Also, I am working on docs for MPC as well as MPPI, will raise a PR soon.