Development of Construction Projects Scheduling with Evolutionary Algorithms

Mehdi Tavakolan

Development of Construction Projects Scheduling with Evolutionary Algorithms
Tavakolan, Mehdi
Thesis Advisor(s):
Betti, Raimondo
Ph.D., Columbia University
Civil Engineering and Engineering Mechanics
Persistent URL:
Evolutionary Algorithms (EAs) as appropriate tools to optimize multi-objective problems have been applied to optimize construction projects in the last two decades. However, studies on improving the convergence ratio and processing time in the most applied algorithms such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) in construction engineering and management domains remain poorly understood. Furthermore, hybrid algorithms such as Hybrid Genetic Algorithm-Particle Swarm Optimization (HGAPSO) and Shuffled Frog Leaping Algorithm (SFLA) have been presented in computational optimization and water resource management domains during recent years to prevent pitfalls of the aforementioned algorithms. In this dissertation, I present three studies on hybrid algorithms to show that our proposed hybrid approaches are superior than existing optimization algorithms in finding better project schedule solutions with less total project cost, shorter total project duration, and less total resources allocation moments. In the first, I present a HGAPSO approach to solve complex, TCRO problems in construction project planning. Our proposed approach uses the fuzzy set theory to characterize uncertainty about the input data (i.e., time, cost, and resources required to perform an activity). In the second, I present the SFLA algorithm to solve TCRO problems using splitting allowed during activities execution. The third study involves the evaluation of the inflation impact on resources unit price during execution of construction projects. This research presents the comprehensive TCRO model by comparing two hybrid algorithms, HGAPSO and SFLA, with the three most capable algorithms -- GA, PSO and ACO -- in six different examples in terms of the structure of projects, construction assumptions and kinds of Time-Cost functions. Each of the three studies helps overcome parts of EAs problems and contributes to obtaining optimal project schedule solutions of total project duration, total project cost and total resources allocation moments of construction projects in the planning stage. The findings have significant implications in improving the scheduling of construction projects.
Civil engineering
Industrial engineering
Item views
text | xml
Suggested Citation:
Mehdi Tavakolan, , Development of Construction Projects Scheduling with Evolutionary Algorithms, Columbia University Academic Commons, .

Columbia University Libraries | Policies | FAQ