A modified genetic algorithm for precedence constrained operation sequencing problem in process planning

Abstract

Precedence constrained operation sequencing problem (PCOSP) is concerned with selection of feasible and efficient operation sequence with minimal machining cost in process planning. Traditional genetic algorithm (GA) generates solution sequence by using randomly selection and insertion of operations, which will break the precedence constraints between operations. The additional fixing approaches for the infeasible solutions will result in low efficiency. Some modified GAs could generate feasible solutions but have premature convergence problem when facing complicated precedence constraints. To overcome the shortcomings, this paper proposed a modified GA that use an edge selection based chromosome encoding approach to make sure all the precedence constraints are met in every step. The experiment illustrates that the proposed GA has superiority in finding optimal or near optimal solution.