Hybrid Evolutionary Algorithms with ML and GPU for Intelligent Scheduling Problems
Fuzzy Logic Systems Institute and Tokyo University of Science, Japan
Abstract: In the real world of scheduling systems, many combinatorial optimization problems (COPs) impose on more complex issues, such as complex structure, nonlinear constraints, and multiple objectives to be handled simultaneously and make the problem intractable to the traditional approaches because of NP-hard COP. In order to develop an efficient solution algorithm that is in a sense “best solution” that is, whose reasonable computational time for NP-hard combinatorial problems met in practice, we have to consider the following very important issues:
- Quality of solution
- Computational time
- Effectiveness of the nondominated solutions
Evolutionary algorithm (EA) is a subset of metaheuristics, a generic population-based metaheuristic such as genetic algorithm (GA), particle swarm optimization (PSO), and estimation of distribution algorithm (EDA). EA is based on principles from evolution theory, and it is very powerful and broadly applicable stochastic search and combinatorial optimization technique which is effective for solving various NP hard COP models.
This tutorial talk will be firstly introduced a brief survey of several metaheuristics based on EA such as hybrid GA (HGA), multiobjective GA (MoGA), PSO and EDA for applying to various combinatorial optimization problems in semiconductor manufacturing systems. Secondly real applications based on hybrid EAs will summarize the following recent scheduling topics:
- Enhancing Hybrid Genetic Algorithms with Machine Learning[14,15]
- HDD Manufacturing Scheduling by HGA and PSO+GA[9,10,12-13]
- TFT-LCD Module Assembly Scheduling by MoHGA with TOPSIS
- Semiconductor Final Testing Scheduling by Cooperative EDA
- Real-time Scheduling for Semiconductor Manufacturing System by MoEDA
- Train Scheduling by MoEA-HSS with Machine Learning & GPU Units
 M. Gen & R. Cheng, 1997: Genetic Algorithms and Engineering Design, 432pp, John Wiley & Sons, New York.
 M. Gen & R. Cheng, 2000: Genetic Algorithms and Engineering Optimization. 512pp, John Wiley & Sons, New York.
 M. Gen, R. Cheng & L. Lin, 2008: Network Models and Optimization: Multiobjective Genetic Algorithm Approach, Springer, London.
 X.J. Yu & M. Gen, 2010: Introduction to Evolutionary Algorithms, 418pp, Springer, London.
 M. Gen, L. Lin & H. Zhang, 2009: “Evolutionary techniques for optimization problems in integrated manufacturing system: State-of-the-art-survey”, Comp. & Ind. Eng., 56(3),779-808.
 M. Gen & L. Lin, 2014: Multiobjective evolutionary algorithm for manufacturing scheduling problems: state-of-the-art survey, J. of Intelligent Manufact., vol. 25, no. 5, pp.849–866.
 C-W Chou, C-F Chien & M. Gen, 2014: A Multiobjective Hybrid Genetic Algorithm for TFT-LCD Module Assembly Scheduling, IEEE Trans. on Automation Sci. and Eng., 10(3) 692-705.
 X-C Hao, J-Z Wu, C-F Chien & M. Gen, 2014: The Cooperative Estimation of Distribution Algorithm: A Novel Approach for Semiconductor Final Test Scheduling Problems, J. of Intelligent Manufacturing, vol.25, no.5, pp.867-879.
 C. Chamnanlor, K. Sethanan, C-F Chien & M. Gen, 2014: Re-entrant flow shop scheduling problem with time windows using hybrid genetic algorithm based on autotuning strategy, International Journal of Production Research, vol.52, no.9, pp.2612-1629.
 M. Gen, L. Lin & W.Q. Zhang, 2015: Multiobjective hybrid genetic algorithms for manufacturing scheduling: Part I Models and Algorithms; Part II Case Studies of HDD and TFT-LCD, Advances in Intelligent Sys. and Comp. 362, 3-52.
 H-K Wang, C-F Chien & M. Gen, 2015: An algorithm of multi-subpopulation parameters with hybrid estimation of Distribution for Semiconductor Scheduling with Constrained Waiting Time, IEEE Trans. on Semiconductor Manufacturing, vol.28, no.3, pp.353-366.
 C. Chamnanlor, K. Sethanan, M. Gen & C-F Chien, 2015: Embedding ant system in genetic algorithm for re-entrant hybrid flow shop scheduling problems with time window constraints, Journal of Intelligent Manufacturing, 17pp. DOI 10.1007/s10845-015-1078-9.
 C. Sangsawang, K. Sethanan, T. Fujimoto & M. Gen, 2015: Metaheuristics optimization approaches for two-stage reentrant flexible flow shop with blocking, Expert Systems with Applications, vol.42, pp.2395–2410.
 M. Gen, W.Q. Zhang, L. Lin, & Y.S. Yun, 2017: Recent Advances in Hybrid Evolutionary Algorithms for Multiobjective Manufacturing Scheduling, Comp. & Ind. Eng., 112, pp.616-633.
 L. Lin & M. Gen, 2018: Hybrid Evolutionary Optimization with Learning for Production Scheduling: State-of-the-Art Survey on Algorithms and Applications, Inter. Journal of Production Research, vol. 56, no.1-2, pp. 193-223.
 K. Nitisiri, M. Gen & H. Ohwada, 2019: A parallel multi-objective genetic algorithm with learning based mutation for railway scheduling, Comp. & Ind. Eng., vol.130, pp.381-394.
Bio-Sketch: Mitsuo Gen – Google Scholar Citation
Dr. Mitsuo Gen is a senior research scientist at Fuzzy Logic Systems Institute and visiting professor at Research Institute for Science and Technology, Tokyo University of Science, Japan.
PhD in Engineering, Kogakuin University in 1975 and PhD in Informatics, Kyoto University in 2006.
Faculty: Ashikaga Institute of Technology 1974-2003 and Waseda University 2003- 2010.
Visiting Faculty: University of California at Berkeley 1999-2000, Texas A&M University 2000, Hanyang University 2010-2012 and National Tsing Hua University 2012-2014, Fuzzy Logic Systems Institute 2009-current and Tokyo University of Science 2014-current.
Research Field: Evolutionary Computation, Manufacturing Scheduling, and Logistics.
Society: President of APIEMS for 2005.1 – 2006.12 and IMS for 2010.8 – 2016.8. Fellow of APIEMS and Honorary member of SOFT. IEEE T-SM (Transactions on Semiconductor Manufacturing) 2015 Best Paper Award in May 2016.