30.06.2013 Views

ดาวน์โหลด All Proceeding - AS Nida

ดาวน์โหลด All Proceeding - AS Nida

ดาวน์โหลด All Proceeding - AS Nida

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Table 2. Brief summary on the recent JSSP articles classifying into four categories.<br />

Category Approaches Authors How Results<br />

Heuristic rules<br />

Mathematical<br />

programming<br />

Neighbourhood<br />

search methods<br />

Artificial<br />

intelligence<br />

Dispatching<br />

rules<br />

Branch and<br />

Bound method<br />

Lagrangian<br />

relaxation<br />

Tabu Search<br />

Simulated<br />

Annealing (SA)<br />

Artificial Neural<br />

Network<br />

Holthaus and<br />

Rajendran [17]<br />

Dabbas et al. [18]<br />

Artigues and<br />

Feillet [3]<br />

Artigues et al. [4]<br />

Baptiste et al. [5]<br />

Chen and Luh [6]<br />

Gröflin and<br />

Klinkert [7]<br />

Zhang and Wu [8]<br />

Yang et al. [19]<br />

Yang and Wang<br />

[20]<br />

Fuzzy logic Bilkay et al. [21]<br />

Petri net<br />

Nakamura et al.<br />

[22]<br />

Presented five new dispatching rules for<br />

scheduling in a job shop<br />

Proposed a modified dispatching rule that<br />

combines multiple dispatching criteria into<br />

a single rule<br />

Proposed a new method based on iterative<br />

solving via branch and bound decisional<br />

versions<br />

Proposed an insertion heuristic and<br />

generalized resource constraint<br />

propagation mechanism.<br />

Compared two Lagrangian relaxations for<br />

the job shop<br />

Developed a new Lagrangian relaxation<br />

approach for JSSP<br />

Integrated the neighbourhood in TS<br />

method and reported quite favourably with<br />

benchmark results<br />

Designed a neighbourhood structure based<br />

on blocks for solving JSSP<br />

Presented an improved constraint<br />

satisfaction adaptive neural network (NN)<br />

for JSSP<br />

Proposed a new hybrid approach,<br />

combining CSANN and two heuristics, for<br />

JSSP<br />

Proposed a fuzzy logic-based algorithm for<br />

assigning priorities to part types that are to<br />

be machined<br />

Proposed an evolutionary method to solve<br />

the problem approximated by generating<br />

the Petri net structure<br />

92<br />

These rules are found to be quite<br />

good in minimizing the mean flow<br />

time<br />

Reduced cycle times and lateness<br />

This method allows to close<br />

previously unsolved benchmark<br />

Results prove the constraint<br />

propagation generalisation when<br />

time lags are considered<br />

These results show that these<br />

problems are greatly intractable<br />

New approach is efficient for large<br />

problems with long time horizons<br />

Solutions improve most of<br />

benchmark result found in this paper<br />

New neighbourhood promotes the<br />

searching capability of SA and helps<br />

it convergence<br />

Outperform the original adaptive NN<br />

in term of computational time and<br />

the quality of schedules<br />

Excellent performance respects to<br />

the quality of solutions and the speed<br />

of calculation.<br />

Proposed algorithm improves the<br />

system efficiency<br />

Experimental evaluation show the<br />

effectiveness of approach

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!