Fuzzy sets and knowledge systems in geotechnical engineering
Abstract
Geotechnical Engineering is a field where exact solutions are guidelines and where analytical models are frequently of limited validity or use. In addition, an important number of relevant parameters are often established in terms of linguistic expressions. Nevertheless, experienced engineers are able to design safe and economical structures by the application of proper engineering judgment, developed from previous experience. This investigation proposes new tools to aid improving the practice of geotechnical engineering. The work begins with the analysis of the geotechnical field from a decision making perspective. Uncertainty is presented as the consequence of subjects' adaptation to their complex environments. Then fuzzy sets and knowledge systems are introduced as potential tools to aid decision makers. Knowledge elicitation and the development of membership functions are discussed at length; experiments are performed to help resolve related issues, and important concepts are formulated. The first application of fuzzy sets is through fuzzy mathematics, in the calculation of the velocity of landslides. Then, a computer system based on fuzzy logic is developed. Concepts formulated in earlier parts of the work are tested, and the system is used to model expert judgment in the analysis of slope stability; an extension of the system to evaluate the safety of existing dams is proposed. The last part of this work concentrates on knowledge systems. A new structure based on fuzzy sets is developed and its potential in artificial intelligence is investigated. IMPROVE, a prototype knowledge system to help decision makers in the area of soil improvement is written with this tool. It incorporates explanation capabilities, combination of alternatives, automatic search for lacunae, relaxation of decisions, and different evaluation functions. The system is capable of suggesting alternatives based on a database of case histories, and runs in a resourceful environment with a built-in geotechnical database, SOIL.
Degree
Ph.D.
Advisors
Chameau, Purdue University.
Subject Area
Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.