Ant colony optimization-based multi-mode scheduling under renewable and nonrenewable resource constraints

a r t i c l e i n f o Keywords: Multi-mode scheduling Ant colony optimization Renewable and nonrenewable resource constraints Construction projects An ant colony optimization (ACO)-based methodology for solving the multi-mode resource-constrained project scheduling problem (MRCPSP) considering both renewable and nonrenewable resources is presented. With regard to the MRCPSP solution consisting of activity sequencing and mode selection, two levels of pheromones are proposed to guide search in the ACO algorithm. Correspondingly, two types of heuristic information and probabilities as well as related calculation algorithms are introduced. Nonrenewable resource-constraint and elitist-rank strategy are taken into account in updating the pheromones. The flowchart of the proposed ACO algorithm is described, where a serial schedule generation scheme is incorporated to transform an ACO solution into a feasible schedule. The parameter-selection and the resultant performance of the proposed ACO methodology are investigated through a series of computational experiments. It is expected to provide an effective alternative methodology for solving the MRCPSP by utilizing the ACO theory. In a construction process, available amounts of renewable and nonrenewable resources may be limited due to cost or other factors. Therefore the resource-constrained project scheduling problem (RCPSP) that has been intensively addressed in the manufacturing field has also been studied in the construction field. In the classical RCPSP, each activity has a single execution mode, that is, the resource requirements and associated duration for an activity are fixed. However, in the practical circumstance when limited resources are shared by multiple activities and the real-time amounts of available resources are thus flexible, some activities may have multiple execution modes to be selected, each mode having different resource amounts corresponding to different durations for each activity. Multiple selections of execution modes are considered to reduce delay of activities and idle times of resources in consideration of real-time amounts of resources. It is shown that the study on the multi-mode resource-constrained project scheduling problem (MRCPSP) is important and has been gradually paid attention in the construction field. The methodologies that have been developed for solving the MRCPSP include the exact and heuristic or meta-heuristic approaches. The exact methods include the exact enumeration schemes [29], a depth-first branch and bound procedure [31], as well as the dynamic programming approaches [27]. In addition, the enumeration scheme [12] for the single-mode was extended to the multi-mode case [32]. Finally, a branch-and-bound algorithm [33] was developed based on the exact procedure [19]. …