Thesis

Master's Thesis (2006)

Artificial Intelligence Paradigms as Decision Support Tools for Real Time Incident Management

The thesis attempts to examine the feasibility of Artificial Intelligence (AI) paradigm as a decision support tool. Case Based Reasoning (CBR) and Support Vector Regression (SVR) have been used to create AI models. PARAMICS has been used to recreate on-site conditions and run simulations. A Section of I- 85, from exit 66 to exit 70 and another section of I-385, from exit 16 to exit 22, have been chosen to represent on-site conditions. VB coding was used generate multiple cases from the study network. All the cases were then simulated to create output training parameter (i.e. time saving rate) for SVR and CBR models. The models were then calibrated. Validation was done using the Vermont network created by Rutgers University.

Bachelor of Engineering Thesis (2003)

Artificial Neural Network (ANN) Models for Strength Prediction of Cement

In this study, an ANN model was developed to predict the strength of cement from physical and chemical parameters. Four different models were developed. The 4 models correlated strength of concrete with physical properties, chemical properties, user friendliness and a combination of physical and chemical properties. The models thus correlated critical characteristics, in all the four parameters, and predicted strength by grading these conditions according to available data. Data regarding physical and chemical parameters was obtained from different cement manufacturing companies. The models were created and trained with the help of ANNS and Qnet (Software to create Artificial Neural Network models) based on this data

Google Sites

Report abuse