A hybrid genetic algorithm for the open shop scheduling problem

In this paper novel hybrid genetic algorithms for Open--Shop Scheduling Problem (OSSP) are presented. Two greedy heuristics LPT-Task and LPT-Machine are proposed for decoding chromosomes represented by permutations with repetitions. For comparison the standard permutation representation of OSSP instances is used. The algorithms apply also efficient crossover operator LOX and mutation operators SWAP and INVERT with constant and variable mutation probabilities. We compare conventional GA to parallel genetic algorithm (PGA) in a migration model. The performance of the algorithms with various settings is verified by computer experiments on a set of large random OSSP instances.