Development of new learning techniques to improve the efficiency of the meta-heuristic algorithms for solving distinct trade-off optimization problems in project scheduling

In this dissertation, new opposition-based learning techniques are developed and incorporated with meta-heuristic algorithms to enhance their performances in order to achieve new and high-quality Pareto-front optimal solutions for the set of trade-off optimization problems in project scheduling. For this purpose, new opposition-based learning (OBL) techniques, namely golden ratio-based opposition learning (GROL), modified dynamic-opposition learning (MDOL), and hybrid opposition learning (HOL), have been developed and combined with meta-heuristic algorithms. Integrating these learning strategies with meta-heuristic algorithms like teaching learning-based optimization (TLBO) and Aquila optimizer (AO) boosts the convergence speed and maintains a fine balance between the global explorative and local exploitative behaviors of the search. The tabular and graphical representation of the numerical simulations reveals that the innovative features arising from the incorporated OBL strategies into the meta-heuristic algorithms played a vital role in reducing the pure randomness of the initial population. Hence, it can be inferred that the proposed models are successful for solving the complex trade-off optimization problems in the construction management domain.