Civil-comp proceedings · 2009 · 12 citations · 5 references
EngineeringCivil EngineeringArtificial IntelligenceBridge EngineeringMachine LearningBridge ConstructionBridge DesignNeural ModelingNeural NetworksArtificial Intelligence In EstimatingConcrete InfrastructureHighway ConstructionHighway EngineeringNeural Network ModelingNeural Network PredictionPrestressed Concrete Bridges
Current models for preliminary quantity estimate of highway bridges are not successful in preparing a quick and economic bill of quantity with sufficient accuracy for bidding purposes. Moreover, most of these models have serious problems in their development and update. In this paper, an effort to develop an Artificial Neural Network (ANN) model that eliminates the shortcomings of the current models is presented. An ANN with back-propagation learning algorithm was trained to estimate the concrete volume and prestressing weight in bridge superstructure. A set of 22 prestressed concrete bridges constructed in Egypt over the Nile was used in training and testing the network. Testing results indicated that ANNs are sufficient tools for the preliminary quantity estimate of highway bridges.
5