Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/923
Title: Artificial intelligence integrated construction simulation method
Authors: Chan, Wah-ho
Keywords: Building -- Computer simulation
Construction industry -- Data processing
Artificial intelligence
Hong Kong Polytechnic University -- Dissertations
Issue Date: 2007
Publisher: The Hong Kong Polytechnic University
Abstract: On one hand, the interactive, complicated system environment of a construction site renders the conventional site layout planning and scheduling techniques inadequate in coping with the materials handling system design in construction. On the other hand, simulation provides a promising solution to construction planning by predicting the future state of a real system from computer model based experiments. However, the complexity and time requirement involved in constructing a simulation model along with the difficulty in decoding simulation output data keep practicing engineers from employing simulation tools in practice. In an attempt to facilitate simulation application in construction, this thesis research has (1) formalized the procedure of setting up a construction operations simulation model in accordance with the Simplified Discrete-Event Simulation Approach (SDESA) and (2) experimented with data mining and knowledge discovery from the data produced by valid construction simulations by applying Artificial Neural Networks (ANN). Implementation and validation of the research findings are based on a real world case for improving the effectiveness of the materials handling system on a precast viaduct construction project in Hong Kong. How to apply the simulation methodology of SDESA is elaborated and illustrated step by step, with the case study. Particular emphasis is placed on the procedures of establishing a simulation model, the validation of the simulation model, the design of simulation experiments, and the analysis of simulation results. With the same case of simulation modeling, this research further demonstrates the application of neural networks (NN)-based sensitivity analysis in support of construction operations simulation modeling. Four input factors relevant to operations and logistics planning in site are identified and assessed, namely (1) the number of trailer trucks rented for hauling segments, (2) one-batch vs. two-batch precast segment delivery modes, (3) the proportion of segments placed in the remote storage area, and (4) the haul duration from the remote storage area to the working span. The NN-generalized input sensitivity information was corroborated by experienced site managers and engineers, which in turn (1) helped validate the operations simulation model, (2) provided valuable insight into the behavior of the operations simulation model, and (3) enhanced the understanding of the real construction system. On the case study project, the field managers were convinced of the functionality and effectiveness of the artificial intelligence integrated construction simulation method being proposed. They took advantage of findings from the research in designing the actual materials handling system. In short, it is hoped that by implementing the simplified, intelligent construction simulation method as developed, practicing engineers would be capable and confident to draw up the best construction plan that would lead to the enhancement of cost-effectiveness and productivity in the field.
Description: vii, 75, [24] p. : ill. ; 30 cm.
PolyU Library Call No.: [THS] LG51 .H577M CSE 2007 ChanW
URI: http://hdl.handle.net/10397/923
Rights: All rights reserved.
Appears in Collections:Thesis

Files in This Item:
File Description SizeFormat 
b21459174_link.htmFor PolyU Users167 BHTMLView/Open
b21459174_ir.pdfFor All Users (Non-printable)1.92 MBAdobe PDFView/Open
Show full item record

Page view(s)

1,078
Last Week
8
Last month
Checked on Jun 18, 2017

Download(s)

12,184
Checked on Jun 18, 2017

Google ScholarTM

Check



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.