Optimizing time–cost in construction projects using modified quasi-opposition learning-based multi-objective Jaya optimizer and multi-criteria decision-making methods

Abstract

This study introduces a modified quasi-opposition learning Jaya optimization (MQOL-Jaya) algorithm to address time–cost-trade-off (TCTP) optimization problems. The proposed method integrates Jaya algorithm with modified quasi-opposite learning (MQOL) during the initial population and generation jumping phases to reduce computational load and enhance solution quality. The effectiveness of the approach is demonstrated on TCTP problems involving 18, 19, and 63 activities. The results reveal that MQOL-Jaya provides competitive solutions, outperforming plain particle swarm optimizaiton (PSO), teaching learning based optimization (TLBO), Jaya, and quasi-oppositional Jaya (QO-Jaya) in terms of function evaluations (NFE), spread (Sp), and hypervolume (HV) indicators. An iterative-based varying weighting factor for MQOL is introduced to improve population diversity and fast convergence. The CRITIC method was used to objectively determine the importance of each criterion, and then the SAW method was used to rank the Pareto front solutions based on these weights. Hence, the basic contribution of this study is MQOL-Jaya approach that provides TCTP resource utilizations (construction plans) to evaluate the impact of these resources on the construction project performance.