Model Predictive Control for a quadrotor in static and dynamic environments

KU Leuven

Optimization of Mechatronic Systems H04U1a

Authors: Elias Rached,Mohamad Sayegh

Overview

Based on the research paper “Model Predictive Control for Aerial Collision Avoidance in dynamic environments”

Goal is to safely navigate in a workspace populated by static and moving obstacles

Objectives

• Implement point-to-point Model Predictive Control

• Obstacle Avoidance in static and dynamic environment (Moving obstacles)

• Finding the best formulation of the problem ( MPC and Obstacles )

• Exploring time optimality (comparison several methods / parameters)

• How to escape Local minima

Application

• Using Rockit: Rockit (Rapid Optimal Control kit) is a software framework to quickly prototype optimal control problems (aka dynamic optimization) that may arise in engineering: iterative learning (ILC), model predictive control (NMPC), motion planning. Link: https://github.com/meco-group/rockit

• Solver used: IPOPT (Interior Point OPTimizer)

• Control input limits: −𝟏 ≤𝑢 ≤𝟏

• OCP horizon length: N = 10 - 20 control intervals

• Sampling time: dt = 0.05 - 0.1s

• Discretization method: multiple shooting

• Numerical integration method: Runge-Kutta 4th order

• Using warmstarting

Dealing with Local minima

Methods of escaping the local minimum are proposed:

• Obstacle merging: when obstacles are close enough they are merged into one obstacle

• A* trajectory guiding: when confirmed to be stuck in a local minimum, an A* trajectory is found, and intermediate points are extracted to guide the drone away

• Potential Field method could help escape certain local minima (if close to obstacles)

Example Results

Dynamic enviroment

Reference paper method with Dynamic Enviroment

Potential Field method with Dynamic Enviroment

Contact

sayeghmohamad@outlook.com