Customized particle swarm optimization for harmonizing multi-section construction projects

Scheduling repetitive construction projects is a challenge due to computational complexity. The authors aim to construct a scheduling system to facilitate the optimization of such projects. First, they formulate a mathematical model of a project composed of any combination of units to be delivered in a fixed order or whose order is to be optimized. With ancillary variables added to the project network, the model proves simple enough to facilitate computations related to schedule optimization. A particle swarm optimization (PSO) algorithm with a unique method to prevent stagnation of the results is proposed to solve the bi-criteria schedule optimization problem. The resulting scheduling system is tested on an actual project to expand an intermodal terminal. The results point to the practical potential of the system. Complemented with a user-friendly interface, it is a ready scheduling support tool. Future research may include incorporating more optimization criteria and making the system riskaware.