Multi-Objective Design Space Exploration for the Integration of Advanced Analytics in Cyber-Physical Production Systems

The integration of advanced data analytics in manufacturing systems has shown impressive results in various fields, including fault diagnosis, predictive maintenance, energy management, and manufacturing system control. However, due to the distributed nature of analytics algorithms and the growing complexity of modern production systems, the performance and the cost of such systems highly depends on the underlying system architecture. Therefore, it is mandatory that system architects systematically explore and evaluate all architectural alternatives of the highly constrained design space defined by the systems functional and economical objectives. This paper presents a design-space-exploration method that not only generates different implementation alternatives, but also provides a formal performance analysis of the generated solutions. By analyzing the architecture of a manufacturing system as well as the data flow graph model of a data analytics algorithm, we automatically allocate, synthesize, and generate different simulatable software solutions to efficiently compute and visualize data analytics algorithms on the shop floor. This approach allows the user to evaluate different architectural implementation during the design phase, to select a solution according to its requirements and to analyze the performance of the resulting system. The applicability of this method is also demonstrated by means of a real world example.

[1]  Wolfgang Rosenstiel,et al.  Integrated requirement evaluation of non-functional system-on-chip properties , 2008, 2008 Forum on Specification, Verification and Design Languages.

[2]  Johan J. Lukkien,et al.  Pattern-based feature extraction for fault detection in quality relevant process control , 2017, 2017 22nd IEEE International Conference on Emerging Technologies and Factory Automation (ETFA).

[3]  C. Mbohwa,et al.  A system model to improve the productivity of a south African steel industry , 2017, 2017 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM).

[4]  Niklas Sörensson,et al.  Translating Pseudo-Boolean Constraints into SAT , 2006, J. Satisf. Boolean Model. Comput..

[5]  Chen-Fu Chien,et al.  A data mining approach for analyzing semiconductor MES and FDC data to enhance overall usage effectiveness (OUE) , 2014, Int. J. Comput. Intell. Syst..

[6]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[7]  Michael Glaß,et al.  A co-simulation approach for control performance analysis during design space exploration of cyber-physical systems , 2011, SIGBED.

[8]  Panagiotis Manolios,et al.  Synthesizing Cyber-Physical Architectural Models with Real-Time Constraints , 2011, CAV.

[9]  Martin Lukasiewycz,et al.  Opt4J: a modular framework for meta-heuristic optimization , 2011, GECCO '11.

[10]  Sudarsan Rachuri,et al.  Predictive Analytics Model for Power Consumption in Manufacturing , 2014 .

[11]  Kalyanmoy Deb,et al.  Simulated Binary Crossover for Continuous Search Space , 1995, Complex Syst..

[12]  Martin Lukasiewycz,et al.  Efficient symbolic multi-objective design space exploration , 2008, 2008 Asia and South Pacific Design Automation Conference.

[13]  Marco Laumanns,et al.  Performance assessment of multiobjective optimizers: an analysis and review , 2003, IEEE Trans. Evol. Comput..

[14]  Petra Hofstedt,et al.  Design space exploration for cyber physical system design using constraint solving , 2013, Proceedings of the 2013 Forum on specification and Design Languages (FDL).

[15]  Enrique Alba,et al.  MOCell: A cellular genetic algorithm for multiobjective optimization , 2009, Int. J. Intell. Syst..

[16]  Andreas Mayr,et al.  Potentials of machine learning in electric drives production using the example of contacting processes and selective magnet assembly , 2017, 2017 7th International Electric Drives Production Conference (EDPC).

[17]  Qing Chang,et al.  Comparison study of two reinforcement learning based real-time control policies for two-machine-one-buffer production system , 2017, 2017 13th IEEE Conference on Automation Science and Engineering (CASE).

[18]  Tuncay Erzurumlu,et al.  Comparison of response surface model with neural network in determining the surface quality of moulded parts , 2007 .