Location

London

Event Website

http://www.csce2016.ca/

Description

Many Researchers have attempted to establish a methodology for the selection of bridge type in a systematic manner. Knowledge based systems (KBS) and other Expert Systems (ES) have been used for this purpose but they have some limitations and restrictions. This paper proposes a methodology to implement a Decision Support System (DSS) in an artificial intelligent environment that aims to suggest a bridge type with its main components at the conceptual design phase, based on the characteristics and performance of existing and similar bridges in order to predict the performance of proposed ones that have been analyzed by decision makers with limited subjectivity. The proposed methodology is divided into three main divisions: 1) this division includes a database that will be structured to store appropriated information besides including models like Point Scale and Quality Function Deployment (QFD) systems that will serve for linguistic conversion to numbers needed for the DSS engine. This division contains as well all the mandatory criteria that have influence on the performance of proposed bridges. 2) this division is the core of the “DSS Engine” where it receives the information from the database that will be implemented in an Artificial Neural Network (ANN) module for training, testing and then predicting the performance of a new case bridge. Afterwards a decision will be made to implement the ANN’s results into a Bridge Information Modeling (BrIM) environment to visualize the suggested design and to predict the potential problems. 3) In this division, a final decision will be made based on the results of the second division. In the proposed DSS, most of the factors are considered as criteria in the database; criteria that have influence on the decision are automatically considered during the analysis process and are introduced in the DSS Engine. The flexibility of the proposed methodology and particularly the database and the method of analysis will make the DSS very helpful in the area of bridge design and management. This will provide bridge engineers with an efficient tool that will minimize the subjectivity in their decisions. A case project will be considered to test the workability and capability of the proposed methodology.


Share

COinS
 
Jun 1st, 12:00 AM Jun 4th, 12:00 AM

STR-996: OPTIMIZATION FOR BRIDGE TYPE SELECTION USING ARTIFICIAL NEURAL NETWORKS

London

Many Researchers have attempted to establish a methodology for the selection of bridge type in a systematic manner. Knowledge based systems (KBS) and other Expert Systems (ES) have been used for this purpose but they have some limitations and restrictions. This paper proposes a methodology to implement a Decision Support System (DSS) in an artificial intelligent environment that aims to suggest a bridge type with its main components at the conceptual design phase, based on the characteristics and performance of existing and similar bridges in order to predict the performance of proposed ones that have been analyzed by decision makers with limited subjectivity. The proposed methodology is divided into three main divisions: 1) this division includes a database that will be structured to store appropriated information besides including models like Point Scale and Quality Function Deployment (QFD) systems that will serve for linguistic conversion to numbers needed for the DSS engine. This division contains as well all the mandatory criteria that have influence on the performance of proposed bridges. 2) this division is the core of the “DSS Engine” where it receives the information from the database that will be implemented in an Artificial Neural Network (ANN) module for training, testing and then predicting the performance of a new case bridge. Afterwards a decision will be made to implement the ANN’s results into a Bridge Information Modeling (BrIM) environment to visualize the suggested design and to predict the potential problems. 3) In this division, a final decision will be made based on the results of the second division. In the proposed DSS, most of the factors are considered as criteria in the database; criteria that have influence on the decision are automatically considered during the analysis process and are introduced in the DSS Engine. The flexibility of the proposed methodology and particularly the database and the method of analysis will make the DSS very helpful in the area of bridge design and management. This will provide bridge engineers with an efficient tool that will minimize the subjectivity in their decisions. A case project will be considered to test the workability and capability of the proposed methodology.

http://ir.lib.uwo.ca/csce2016/London/Structural/117