Contents I. INTRODUCTION
Due to the complexity of the FJSP, no exact method has so far been developed to be able to tackle the problem within a reasonable amount of time (Bagheri et al. 2010) [1]. In recently years, many heuristics have been developed to solve the FJSPs, such as the tabu search (TS) (Brandimarte (1993) [2], Mastrolilli et al. (2000) [3], Saddi-mehrabad et al. (2007) [4], Fattahi et al. (2007) [5], Ennigrou et al. (2008) [6], Li et al. (2010) [7], and Bozejko et al. (2010) [8]), the genetic algorithm (GA) (Kacem et al. (2002) [9], Ho et al (2007) [10], Pezzella et al. (2008) [11] and Gao et al. (2008) [12]), the particle swarm optimization (PSO) (Xia et al. (2005) [13], Gao et al. (2006) [14], Liu et al. (2007) [15] and Zhang et al. (2009) [16]). Also, some very recently published literature reports other meta-heuristics for the problem, such as the parallel variable neighborhood search (PVNS) algorithm developed by Yazdani et al. (2010) [17], the knowledge-based ant colony optimization (KBACO) algorithm proposed by Xing et al. (2010) [18], the artificial immune algorithm (AIA) investigated by Bagheri et al. (2010) [1], and the climbing depth-bounded discrepancy search (CDDS) algorithm (Hmida et al. 2010) [19] are among the most efficient results set. For solving the multi-objective FJSPs, Kacem et al. (2002) [20] and Moslehi et al. (2010) [21] proposed a Pareto based approach which can provide non-dominated solutions. The present multi-objectives are focused on the minimization of the maximum completion time (makespan), the total workload of all machines, and the maximum workload. There are very few literature considers the total flow time criterion for the FJSPs.
