The Civil Engineering Department at the College of Engineering, University of Baghdad, held Ph. D thesis examination titled:
Integrated Model for parametric Construction Cost Prediction using
Artificial Intelligence Techniques)
The examining committee was:
- Prof.Dr.Sawsan Rasheed chairman
- Prof.Dr. Hatem Khalifa member
- Prof. Dr.Waleed Mustafa KHammas member
- ا Assist. Prof. Dr. Mervet Razak Walli member
- Assist. Prof.Dr. Ahmmed M. Raoof member
- Assist. Prof.Dr. Sedqi Ismael Razoki supervisor
Integrated Model for parametric Construction Cost Prediction using
Artificial Intelligence Techniques
Parametric costs are standards that develop the cost of a construction project based on examining and verifying the relationships between construction projects in terms of technical parameters, cost, time and quality involved in the construction project. Parameters can be classified as simple or complex, therefore may constitute a source of concern for the owner of the project or the stakeholders alike because of the difficulty of controlling it.
The main objective of this thesis is to build and analyze an integrated parametric cost management model, which in turn adopts a program to find out the sensitivity of parametric cost to construction projects with values and concepts through a hybrid conceptual framework that means discovering problems arising from risks and complexities in which it is difficult to determine the standard cost accurately through iterative tests and continuous improvements, bridging the gap between an integrated parametric cost program and traditional cost management concepts, and developing a framework for managing construction project cost standards using a computer program (IMPs) Integrated Management Program system. The secondary objectives of this thesis is to conduct an analytical study of the integrated parametric cost program through a case study.
79 parameters of the parametric cost affecting the construction projects were studied using (FAHP,FDM,DELPHI METHOD) technology in order to ensure the accuracy of the probability and impact of the analysis of the parameters on the construction projects. The remaining parameters are also influential, but not applied in Iraq. The most parameters represent these factors as independent inputs to the models used in prediction.
By surveying the construction projects to include them and registering 100 construction projects as a case study and using a measurement form to calculate the impact of parametric cost from the data of construction projects and calculate them in a quantitative or qualitative way and according to the type of influencing standard as well as extracting the correlation for it that for the purpose of preparing it to build five integrated models for predicting which were On two types (supervision and unsupervised), data were trained for ten projects to verify and evaluate the models that were built in the models, which are the supervision models, whose data are defined as inputs by a computer technology (k-means algorithm ,MFGA, Bayesian model), and an accuracy factor (1.9025, 0.0017, 0.49) successively and (unsupervised) models whose data is defined in the form of derivation of equations, which are (stochastic, deterministic), and an accuracy rate (High, Ro<1) was obtained successively, and it was concluded that all five prediction models that were built have the possibility of prediction In standard costs and over varying short, medium and long ranges.
In order to find the sensitivity of parametric costs on construction projects, the results of the five models used were combined with sensitivity analysis algorithms in multiple scenarios to find out their impact on construction projects at different levels in order to help decision makers predict parametric costs in the initial stages of the project. Finally, many conclusions were reached. And recommendations, in addition to suggesting a number of future studies.
After discussing and hearing to the defense of the student the thesis was accepted with excellent degree.
