You will be embedded in the MECO (Motion Estimation Control and Optimization) research team of the KU Leuven Department of Mechanical Engineering. The MECO team focusses on the identification, analysis and control of mechatronic systems such as autonomous guided vehicles, robots, and machine tools. It combines theoretical innovations with experimental validations. The theoretical research benefits from the team’s expertise on numerical optimization, while MECO’s practical knowhow and industrial collaboration are supported by its participation in Flanders Make - a strategic research center for the manufacturing industry.
In the realm of Industry 4.0, mechatronic systems such as autonomous guided vehicles (AGVs), drones, robots, and machine tools are getting more modular and collaborative. This transition drives a paradigm shift in the corresponding learning and control algorithms as well, where centralized single-agent approaches are making way for collaborative, multi-agent counterparts. In addition, optimization-based learning and control (MPC) approaches are steadily gaining ground on classical PID type approaches. Under the hood, modern multi-agent learning and control approaches rely on distributed optimization algorithms to decompose the overall problem in single-agent sub-problems, which are gradually adjusted to yield the overall solution through inter-agent communication.
In this research you will tailor distributed optimization algorithms to multi-agent learning and model predictive control (MPC) approaches. The aim of multi-agent learning is to collaboratively learn optimal control actions (feedforward) from past experience of all agents / subsystems, whereas multi-agent model predictive control aims at collaboratively and optimally adjusting the controls based on real-time sensory information (feedback). The primary application domain of your methods will be motion planning and control for mechatronic systems, particularly AGV’s and mobile robotic manipulators. Your research will mainly focus on software developments, simulation based evaluation and experimental validation on lab-scale setups. For the software developments, you will extend recent developments within the MECO research team on optimal control (check out https://gitlab.mech.kuleuven.be/meco-software/rockit) and learning control (check out https://gitlab.mech.kuleuven.be/meco-software/rofalt). Under the hood, CasADi (www.casadi.org) isused as symbolic optimization framework and algorithmic differentiation tool.
Ideal candidates hold a Master’s degree in engineering, computer science, or applied mathematics. Successful candidates are typically ranked at or near the top of their classes, have a solid background in optimization, systems and control, relevant computer programming skills (Python or Matlab, C++), and enthusiasm for scientific research. Team player mentality, independence, and problem solving attitude are expected, and proficiency in English is a requirement.
Applicants whose mother tongue is neither Dutch nor English must present an official language test report. The acceptable tests are TOEFL, IELTS, and Cambridge Certificate in Advanced English (CAE) or Cambridge Certificate of Proficiency in English (CPE). Required minimum scores are:
A fully funded PhD position in an international context for four years at the KULeuven: a top European university and a hub for interdisciplinary research in the fields of systems, control and optimization. You will be embedded in the MECO research team of the Department of Mechanical Engineering. The doctoral candidate will work in world-class facilities with highly qualified experts, and will benefit from the training scheme developed based on the expertise of academic and industrial partners. A start date in the course of 2020 is to be agreed upon.
Please use the online application tool to submit your application and include:
 an academic CV with photo,
 a Pdf of your diplomas and transcript of course work and grades,
 statement of research interests and career goals (max. 2 pages),
 sample of technical writing (publication or thesis),
 contact details of at least two referees,
 proof of English language proficiency test results.
You can apply for this job no later than February 29, 2020 via the online application tool
KU Leuven seeks to foster an environment where all talents can flourish, regardless of gender, age, cultural background, nationality or impairments. If you have any questions relating to accessibility or support, please contact us at diversiteit.HR@kuleuven.be.Mehr
|Titel||Multi-agent Learning and Control for Mechatronic Systems|
|Job location||Oude Markt 13, 3000 Leuven|
|Veröffentlicht||November 19, 2019|
|Bewerbungsschluss||Februar 29, 2020|
|Jobart||PhD/ Doktorand/in  |
|Fachbereiche||Algorithmen,   Mechatronik,   Programmiersprachen,   Steuerungstechnik,   Robotertechnik,   Angewandte Mathematik,   Maschinenbau,   Computergestütze Mathematik,   Maschinelles Lernen,   |