چكيده به لاتين
Nowadays, most of the construction projects are delayed due to a single objective, or end with different cost and quality what is expected to be. On the other hand, due to the competitive conditions governing the contracting companies, today paying attention to several objectives simultaneously has become a vital issue in the construction industry. Therefore, the project managers are always faced with challenges to create optimal balance between different objectives and often in conflict of project (multi-objective optimization of project scheduling).
In this study, in order to optimize the multi-objective project scheduling, simulation by MATLAB (2016) is performed using fuzzy logic combination and multi-objective genetic algorithm in addition to solving the multi-objective problem and reaching the Pareto front, a practical step is taken to make the proposed multi-objective model more intelligent and make it more like reality. In fact, by fuzzy adaptive system embedding in multi-objective genetic algorithm (FACS model) in an innovative and effective way, the problems of scheduling construction projects are solved.
The results of the proposed FACS model were compared with previous models for similar construction problems and it was observed that while the FACS genetic algorithm was performing properly in optimizing all three project objectives, the evolutionary process of the genetic algorithm was also closer to reality due to the variability of the crossover rate and its dependence on the answers of previous generations. On the other hand, the third objective of the project (project quality), which is indirectly maximized by the fuzzy adaptive system, has provided acceptable answers compared to the previous three-objective models. In fact, the effective and practical performance of the fuzzy adaptive system applied, both in terms of reducing computational complexity and in the quality and diversity of the final Pareto responses, is shown in this study. For example, the solution presented by FACS for the first case study in 59 days at a cost of approximately 155,000 has achieved a quality of over 93%; whereas the solution presented by Afshar et al.'s three-objective model has optimized all three of the project's objectives more poorly than FACS.
