Resource Utilization Optimization using Genetic Algorithm based on Variation of Resource Fluctuation Moment for Extra-Large Building Renovation

Abstract

This paper compared optimizing resource utilization using Genetic Algorithm (GA) based on variation of resource fluctuation moment (Mx) for extra-large building renovation. The Mx variable, as determined by resource demand on day squared, has a large effect on the result of multi-objective optimization. In this research, an Mx-based optimization model targeting five variables for resource utilization was proposed. In addition, the proposed method is flexible in that the construction planners could specify predecessors or preferences for optional construction sequences so that a more efficient optimal scheduling may be obtained. Three activation functions for Mx were considered, namely Mx, and Mx/1000. In this work, the models in consideration were applied to real data from the university main library building renovation projects which consisted of 251 activities. The contractor's work plan was used as the initial scheduling for the optimization process. When comparing the experimental results from all 3 models, it can be seen that the form and Mx/1000 are more suitable in optimizing resource utilization through GA method in extra-large building renovation.