CRAN - Campus Sciences
BP 70239 - 54506 VANDOEUVRE Cedex
Tél : +33 (0)3 72 74 52 90
Ph. D. Project : Hybrid control based on simulation to deal with the unexpected in Manufacturing and Logistics Systems
Dates : 2017/10/01 - 2020/09/30
Manager(s) CRAN: André THOMAS , Hind BRIL
Full reference: In order to provide overall efficient production performance and ensure reactivity face to unpredicted events, the industrial experiences and research activities have demonstrate the interest of a hybrid control system that couple a predictive centralised mode when the manufacturing schedules are generated based on optimization or heuristic algorithm with a distributed mode based on autonomous entities that consider every event in real time, with no anticipation (human-operator or holon/agent).

In such hybrid architectures, the fundamental decision facing perturbation is whether to still follow the predictive/proactive schedule (predictive mode) or not. If not, they may switch to a reactive mode where events and decision are handled in real time with the intention to switch back to a predictive mode as soon as possible. The main issue for researchers is then to provide accurate mechanisms to define the best switching dates (and/or the best switching decision-making levels) for control holons/agents so that they behave in a sense that the behaviour of the hybrid architecture stay globally optimized despite disturbances.

We believe that the use of a dynamic data driven simulation approach as a core of the hybrid control system is a promising way to deal with the switching decision process. The principle is that the simulation system is continually influenced by real time data coming from intelligent objects (products, resources, human…) for better analysis and prediction.

The use of dynamic simulation induces several challenges: The simulation model has to work on real-time mode to ensure the coherence between the proposed solution and the state of the physical system and to work on discrete-events mode to accelerate the simulation. The second challenge is about the simulation model adaptability faced to physical system and environment changes. The simulation application would have in order to support “hot swapping” by inclusion of new data during execution time without stopping the system. In the simulation modelling process, certain phases (model validation, scenarios definition) seem to be entirely under the
responsibility of the human expert. The simulation model design must take into account be human and simulation application interactions.

Thus, the interested applicant must justify good skills on simulation methods and tools, on manufacturing control system and system engineering (human centric approach).
Eco-Technic systems engineering
Financial aspects: Fellowship from research ministry
Publications: hal-00854840v1,hal-01182909    + CRAN - Publications