Lower-level decision task solution while optimising a construction project schedule

Abstract

A bi-level approach for optimisation of a construction project is discussed in the paper. Considered optimisation problem deals with identification of optimal project and a corresponding optimal schedule. Project structure is defined by applied order of technological operations. Application of decomposition-coordination principle facilitates problem solution. We therefore obtain tasks which belong to 2 distinct optimisation levels. The lower optimisation task level is devoted to optimal allocation of execution modes to operations while assuming considered project structures. The global optimisation task level pertains to choice of the best structure for a construction project. Solution of lower level tasks are applied in this regard. The main difficulty in global project schedule optimisation results from multiplicity of feasible construction project structures and a need for solution of lower level tasks. We generally consider application of Monte Carlo simulation (MC) for generating feasible project structures. Mixed linear programming (MILP) and MC is the applied to solve lower level tasks. We also apply metaheurstics combined with MILP to solve lower level tasks while generating feasible project structures. Effects of application of presented approaches for solving lower level task solution approaches are finally compared.