Time–cost trade-off optimization in generalized construction projects using an opposition learning-augmented multi-objective Jaya algorithm

Abstract

This study introduces a multi-objective Jaya (MOO-Jaya) algorithm to unravel time–cost trade-off problems (TCTPs) in construction project scheduling. The model integrates opposition-based learning (OBL) to enhance population initialization and generation jumping mechanisms, thereby improving solution diversity and convergence efficiency. To evaluate performance, the MOO-Jaya algorithm is tested on a real-world construction project comprising 29 activities with complex precedence constraints. The project accounts for generalized precedence relationships (GPRs), including start-to-start (SS), start-to-finish (SF), finish-to-start (FS), and finish-to-finish (FF) activity dependencies, with both positive and negative lag times, enabling realistic modeling of activity overlapping and schedule compression. Computational results are benchmarked against established metaheuristics like the non-dominated sorting genetic algorithm II (NSGA-II), hybrid genetic algorithm with quantum simulated annealing (HGAQSA), and the core Jaya algorithm. The suggested algorithm demonstrates superior Pareto front convergence, solution diversity, and computational efficiency compared to these counterparts. Findings underscore its practical applicability in addressing multi-criteria decision-making problems, offering project planners a robust tool for optimizing time and cost objectives under complex scheduling constraints.