Code for the Implementation of State-Dependent dynamic tube model predictive control

The code of this repository is used to implement and test a new algorithm called SDD-TMPC. It is a control algorithm that sacrifices a bit of optimality (much less than tube MPC) but returns robust solutions. In this repository, MPC, TMPC, and SDD-TMPC were implemented.

SDD-TMPC needs to have a model of boundaries of future disturbance. I used a fuzzy logic-based model trained with a genetic algorithm.

4 scenarios were created to compare all methods:

1. Following a path in an empty trajectory

2. Moving closely to the wall

3. Moving through a narrow corridor.

4. Avoiding an obstacle