A genetic local search algorithm for minimizing total weighted tardiness in the job-shop scheduling problem

Weighted Tardiness Minimization in Job Shops with Setup Times by Hybrid Genetic Algorithm

“…Other authors, for example Mati et al in [7] or Essafi et al in [3] opted to use more accurate estimations (they report results with exact estimations from 57% to 76%, depending on the particular instance). However, their estimation procedure is more time consuming as the complexity goes up from O(1) to O(N ) for each path, where N is the number of jobs.…”

“…Firstly, we compare our algorithm with the GLS algorithm proposed in [3] and the M DL algorithm proposed in [7] to solve the JSP. We experimented across the 22 instances of size 10×10 proposed by Singer and Pinedo in [9] (ABZ5, ABZ6, LA16 to LA24, MT10, and ORB01 to ORB10).…”

“…They propose a shifting bottleneck algorithm and in their experimental study this algorithm is compared with very simple heuristics (some priority rules) across randomly generated instances that are not available for further comparison. However, weighted tardiness has been largely considered for the classic JSP, maybe the algorithms proposed in [3] and [7] being the most relevant approaches currently. In [3], Essafi et al propose a hybrid genetic algorithm that uses a local search based in reversing critical arcs and an algorithm that iterates between hill climbing and random generation of neighbors to escape from local optima.…”

“…However, weighted tardiness has been largely considered for the classic JSP, maybe the algorithms proposed in [3] and [7] being the most relevant approaches currently. In [3], Essafi et al propose a hybrid genetic algorithm that uses a local search based in reversing critical arcs and an algorithm that iterates between hill climbing and random generation of neighbors to escape from local optima. In [7] Mati et al propose a local search method that uses estimators to evaluate the neighbors and that is capable of minimizing any regular objective function, i.e.…”

Metaheuristics for the multi-objective FJSP with sequence-dependent set-up times, auxiliary resources and machine down time

“…Processing times in the JSSP with batching is determined by the number of similar products which are produced simultaneously as part of the same operation (Petrovic et al 2005;Potts et al 1998;and Potts and Kovalyov 2000). Minimizing a function of job due dates, as in the JSSP with due dates, indirectly places restrictions on the processing times of jobs (Essafi et al 2008 andSinger andPinedo 1998). Operation starting and completion times in an expanded JSSP are restricted by release dates, due dates and technological enabling constraints Liang 2001 andZhao et al 2005).…”