Title: GenomIc Model prEdictive ConTrol Tools for evolutionAry pLants
Area: 2- Evolutionary and Reconfigurable Factory
Call topic: 2.1 – Optimization of the co-evolution of production systems
Coordinating Institute: CNR, Institute of Electronics, Computer and Telecommunication Engineering (IEIIT)
- CNR, Institute of Industrial Technologies and Automation (ITIA)
- Politecnico di Milano, Dipartimento di Elettronica, Informazione e Bioingegneria (DEIB)
Industrial Interest Group:
- Fimi (San Giuliano Milanese (MI)
- Prometeo S.r.l. (Montelupo Fiorentino (FI)
- FF di Falconi V. & C. snc (Montirone (BS))
- DPI s.r.l. (Cassano d’Adda (MI))
Abstract: The IMET2AL project proposes a genomic Model Predictive Control (MPC) based software prototype tool that supports the industrial engineers to study and design control system configurations for automated factory production systems characterized by a fast evolutionary behaviour. The obtained control solutions are optimized on the base of key performance indexes like flow production, peak of the absorbed electrical power and the total energy consumed by the plant and they are able to impress to the production system the desired functional behaviour.
The platform tool is structured in two level control structure where, at the lower layer, distributed MPC algorithms will be used to control the individual equipments of the factory production system while at the upper layer an MPC coordinator will be designed by taking fully advantage of the most recent advances in hybrid control theory, dynamic programming, mixed‐integer optimization, and game theory.
There are two major benefits coming from the IMET2AL project.
The first and most important one regards the sharing and the development among the project partners of the knowledge on advanced model predictive control techniques to be applied to manufacturing industrial plants. These methodologies are typically used in the process industry but not in the manufacturing one and will open the manufacturing industry to new production system scenarios characterized by optimized and high efficiency performances.
The second benefit concerns the development of a prototype software platform tool useful to support the industrial production system engineer in designing the plant automation system based on an abstract genomic control model so as to characterize the whole automated factory production system according to its evolutionary behaviour.
Ideas and solutions:
- The tool is structured in two layers. At the lower layer, distributed MPC algorithms control individual equipment of the factory production system. At the upper layer an MPC coordinator takes full advantage of the most recent advances in hybrid control theory, dynamic programming, mixed‐integer optimization, and game theory.
- Sharing and development among project partners of knowledge on advanced model predictive control techniques, and their application to manufacturing industrial plants. These methodologies are not yet widely used in manufacturing and enable the manufacturing industry to new scenarios characterized by optimized and highly efficient
- Development of a prototype software platform tool to support industrial production system engineers in designing the plant automation system so as to characterize the whole automated factory production system according to its evolutionary behavior.
- Further progress with respect to the state of the art can be expected for what concerns flexible control kernels execution, whose range of application will probably extend well beyond the scope of the project. The execution and communication environment to be developed in the project can profitably be adopted whenever flexible, configurable, and platform‐independent deployment of real‐time and best‐effort software modules is needed, within the context of a complex industrial plant.
- The benefits brought by the IMET2AL project has been demonstrated by means of specific tests carried out on a hybrid simulation model of a re‐manufacturing plant, for what concerns the upper level of the prototype software platform. Regarding the lower level, a specific operating machine (automated reworking machine) belonging to a re‐manufacturing plant has been considered and then modelled in order to test the control kernels.
- Virtualization techniques are already very common in office automation and data centers. In the context of the IMET2AL project, the ability to host and execute control kernels within virtual machines allows the designer to think about the control system in terms of its high-level features and functions, rather than how they will be practically deployed and realized at run time. In turn, project outcomes will likely further facilitate the adoption of virtualization techniques for industrial control applications.