Network resource allocation with support of a fuzzy expert system

Critical path method and program evaluation and review technique (CPM/PERT) without resource allocation are totally inadequate for control of major projects. Resource allocation has been a major concern of research and actual application development since the advent of the CPM/PERT network scheduling techniques. Many algorithms have been developed including theoretical optimization and heuristic rules since early 1960. The theoretical approaches generate too many combinations and are, therefore, quite expensive to use. The heuristic approaches provide a reasonable solution to the “too expensive” problem. However, none of these algorithms has the capability to resolve the impacts of a variety of external forces, such as weather conditions and resource availability, that are unavoidable in any real-life project. This paper proposes a practical algorithm to implement resource allocation. The proposed algorithm uses the output from an expert system called priority ranking, which uses possibility theory and the generalized modus ponens logic inference rule to determine the priority ranking. In addition, both network-related and project-related factors are included in the algorithm.