Automated Construction‐Simulation Optimization

Abstract

This paper discusses the basis for automating system optimization through simulation‐feedback analysis. Optimizing site‐management objectives can be attained by manipulating resources in a construction operation provided that the operation is represented by a simulation model (e.g. CYCLONE, SLAMSYSTEM, or GPSS). Optimization can be achieved by iteratively trying different resource‐allocation schemes. To avoid an exhaustive enumeration and trial of all possible combinations in a given system, the paper presents techniques that are useful in guiding the system in its search for the most appropriate allocation. Starting with an arbitrary resource combination, a pilot‐simulation run is performed. Depending on the specified objectives, resource performance is analyzed and a new combination is recommended to get closer to the desired optimum allocation. Three commonly encountered objectives in construction simulation are discussed and optimizing‐system performance using the proposed method to attain these objectives is presented. The particular scenarios covered are minimizing unit cost, maximizing production rate, and reasonable matching among resources.