A three-stage solution algorithm for chemical production scheduling

A novel algorithm that exploits the strengths of discrete- and continuous-time scheduling mathematical programming formulations is proposed. It consists of three stages, in which (i) an approximate solution is obtained using a discrete-time mixed-integer programming scheduling model, (ii) the solution is mapped onto continuous-time grids via a mapping algorithm, and (iii) the accuracy of the solution is improved by solving a continuous-time linear programming model. Two types of continuous-time grids are introduced, namely unit- and material-specific grids, to ensure feasible unit utilization and material balance in the model. An extensive computational study is performed, showing that the algorithm is capable of finding high quality solutions orders of magnitudes faster than traditional methods.