Burak Yüksel

I am a guest scientist in the group at the Max Planck Institute for Biological Cybernetics.

I currently work as a development engineer @ITK-Engineering GmbH (100% subsidiary of Bosch GmbH).

I have completed my PhD on physical interaction of flying robots with in the group at the Max Planck Institute for Biological Cybernetics. I have defended my thesis on 17.02.2017 at IST-University of Stuttgart (Prof. Allgöwer) with magna cum laude.

You can find my personal webpages here:

My research adresses the problem of safe physical interaction of quadrotor UAVs using energy shaping and passivity based controllers. Moreover, I study desing and control of flying robots equipped with rigid or elastic manipulator arms.

I have worked on the  in  at Max Planck Institute for biological Cybernetics.

I have background on:

Robotics:

- Multi-rotor flying robots
- Manipulators
- Multi rgid/elastic body kinematics and dynamics.
- Bipeds, Passive dynamic walking. ( Master Thesis )

Control Theory and Optimization:

- Linear/ Nonlinear Control Theory:
+  Differential flattness, exact linearization
+  Passivity-based control
+  Sliding-mode, back-stepping control
- Optimal Control Theory
- Dynamic Programming    (Visiting Scholar @ Carnegie Mellon University)
- Robust Control

Automotive

- Plug-in Hybrid Electric Vehicles  (Visiting Scholar @ Carnegie Mellon University)  
- Solar powered race cars
- Tilting three wheelers  ( Bachelor Thesis )

Design, Modeling and Control of Aerial Robots for Physical Interaction and Manipulation

Introduction

The physical interaction of flying machines, a.k.a. aerial physical interaction, means that a flying machine (or robot) can exert desired forces and torques to its environment while it can preserve its stable flight. Henc e it is today in interest of many scientists, in sense of developing new control algorithms and designing novel interaction tools .

Goal

The goal of this work is to make a quadrotor, a four propeller light-weight flying robot, physically interact with its environment, while it maintains a stable flight.

Approaches

We tackle the existing problems to achieve the project goal in three different ways.

1. Studying nonlinear system and control theory for aerial physical interaction:
   
   We are developing new control algorithms for aerial physical interaction. It is necessary to describe the problem/goal at hand in a meaningful way. For this reason, we first presented Interconnection and Damping Assignment Passivity Based Control (IDA-PBC) framework for quadrotors in . The idea is to have a controller which shapes the physical properties of a quadrotor for achieving a desired interaction behavior. In parallel to this work, we also study on different nonlinear control theories, and the results will be shared here once they are published.
   
   
2. Estimation theory:
   
   We develop estimation algorithms for interaction force and torques acting on a quadrotor robot. Since force and torque measurement is an expensive procedure in sense of power, money and weight; we introduced a nonlinear force observer for quadrotors in .
   
   
3. Design and manufacturing of novel tools and hardware/software development
   
   We have designed a novel light-weight manipulator for aerial physical interaction (see Fig.4), which can easily equipped with a small-sized aerial vehicles. More particularly, it is designed to be an elastic actuator, which is known to be providing passive compliance, which is very important for stable physical interaction. For this reason, we have developed a flexible-joint arm for flying robots such as quadrotors, to improve their aerial physical interaction capabilities .
   
   Furthermore, we are working on low-level control algorithms for brushless controllers, flight controllers and small-size computers for not only making aerial physical interaction possible, but also achieving better flying performance. To have an idea on what hardware/software we are using, see .
   
   

Ongoing works:

Our current effort focuses on the following topics: