A discrete Water Wave Optimization algorithm for no-wait flow shop scheduling problem. (January 2018)

• Record Type: Journal Article
• Authors: Zhao, Fuqing; Liu, Huan; Zhang, Yi; Ma, Weimin; Zhang, Chuck
• Is Part Of: Expert systems with applications. Volume 91(2018)
• Page Start: 347
• Page End: 363
• Publication Date: 2018-01
• Subjects: Water Wave Optimization (WWO) -- Iterated greedy algorithm -- No-wait flow shop scheduling problem -- Makespan; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2017.09.028
• Languages: English
• ISSNs: 0957-4174

Abstract

Highlights: A Discrete Water Wave Optimization (DWWO) Algorithm is proposed. An Improved Iterated greedy algorithm is integrated into the framework of DWWO. A modified initialization strategy is proposed to generate the initial population. A ruling out inferior solution mechanism is added to improve the convergence speed. The convergence of the DWWO algorithm has been proved theoretically. Abstract: In this paper, a discrete Water Wave Optimization algorithm (DWWO) is proposed to solve the no-wait flowshop scheduling problem (NWFSP) with respect to the makespan criterion. Inspired by the shallow water wave theory, the original Water Wave Optimization (WWO) is constructed for global optimization problems with propagation, refraction and breaking operators. The operators to adapt to the combinatorial optimization problems are redefined. A dynamic iterated greedy algorithm with a changing removing size is employed as the propagation operator to enhance the exploration ability. In refraction operator, a crossover strategy is employed by DWWO to avoid the algorithm falling into local optima. To improve the exploitation ability of local search, an insertion-based local search scheme which is utilized as breaking operator, is applied to search for a better solution around the current optimal solution. A ruling out inferior solution operator is also introduced to improve the convergence speed. The global convergence performance of the DWWO is analyzed with the Markov model. In addition, the computational results based on well-known benchmarks and statistical performance comparisons are presented. Experimental results demonstrate the effectiveness and efficiency of the proposed DWWO algorithm for solving NWFSP.