Optimization-Based Reference Generator for Nonlinear MPC

Project Overview

Model Predictive Control (MPC) approaches often rely on heuristic references and careful tuning. This project introduces an optimization-based reference generator that leverages a Linear Inverted Pendulum (LIP) model to calculate reference trajectories for the center of mass.

Key benefits include:

• Automated Tuning: Eliminates the need for manual tuning of calculating reference trajectories.
• Improved Disturbance Rejection: Enhances performance when encountering environmental disturbances.
• Validated Performance: Tested through simulations and experiments on the Aliengo robot.