A large step random walk for minimizing total weighted tardiness in a job shop

Scheduling Abstractions for Local Search

“…The table reports the optimal value (O), the number of times CT finds the optimum (#O), the best, average, and worst values found by CT, as well as the average CPU time to the best solution. The table also reproduces the result given in [3], which only reports the average of LSRW over 5 runs and the number of times the optimum was found.…”

“…We used the value 1.3 for f which produces the hardest problems in [3] and ran each benchmark 50 times. The table reports the optimal value (O), the number of times CT finds the optimum (#O), the best, average, and worst values found by CT, as well as the average CPU time to the best solution.…”

A hybrid shifting bottleneck-tabu search heuristic for the job shop total weighted tardiness problem

“…The optimal solutions for these instances are obtained by Singer and Pinedo [4]; however, it appears that in this paper the branch-and-bound algorithm was either stopped prematurely or the due dates were inadvertently set too tight because solutions better than those reported by Singer and Pinedo [4] appear in subsequent research. Kreipl [8], De Bontridder [9], and Essafi et al [10] all demonstrate the quality of their algorithms on this set of instances, and we follow suit. In addition, Essafi et al [10] create a new set of benchmark instances for JS-TWT based on the instances created by Lawrence [23] and frequently used for Jm//C max .…”

Minimizing weighted tardiness of job-shop scheduling using a hybrid genetic algorithm

“…In addition, we also did a series of comparisons between the hybrid GA and the heuristic methods (LSRW(15) and LSRW (200))-referred to herein as SPK-proposed in Kreipl (2000) and based on Singer and Pinedo (1998) and Pinedo and Singer (1999). We take the 22 cases used in Kreipl (2000) (abz5, abz6, la16, la17, la18, la19, la20, la21, la22, la23, la24, mt10, orb1, orb2, orb3, orb4, orb5, orb6, orb7, orb8, orb9 and orb10) from the OR library for comparison.…”

“…Further, lead-time iterations are applicable only to heuristics involving due dates or slack time, whereas our hybrid algorithm can be applied to all kinds of heuristics. As a solution procedure for Jmk P T j and Jmk P w j T j , the particular combination of GA-R&M is found to be considerably superior to many popular heuristics (improved with lead-time iterations), such as SPT, EDD, EODD (earliest operation due date), MST (minimum slack time), WLWKR (weighted least work remaining), weighted Hodgson, weighted COVERT and R&M. However, it is not as good as the method by Kreipl (2000), a method specifically designed for minimizing weighted tardiness in a job-shop. The hybrid GA outperforms the GA in Mattfeld and Bierwirth (2004) in a majority of the test cases.…”