A local constraint based analysis approach to project scheduling under general resource constraints

In the doubly-constrained project scheduling problem where nonrenewable resources are constrained both on a period and a project basis, the combinatorial complexity arising from interrelated precedence, renewable and nonrenewable resource constraints necessitates a solution approach which takes all such considerations into account including the objectives. In this context an activity is permitted to operate in one of its modes, each of which represents the trade off between different choices of resource requirements (types) and operating durations. The proposed heuristic approach to handle this problem is local constraint based analysis (LCBA) where the selection of activities and their respective modes is made locally at every decision point. LCBA imposes precedence relationships on activities which guarantee the feasibility of the current makespan constrained by the available resource levels at that decision point as long as a complete sequence of schedulable activities is found. The procedure is also adapted to various model extensions such as flexible resource requirement levels. The procedure can also be utilized in a dynamic environment where resource absenteeism, activity duration changes and readjustment of the project network configuration are accounted for. Numerical results are obtained for LCBA and three well-reputed dispatching rules to show that considerably better solutions are obtained by the LCBA approach in encouragingly short computation times.