Knowledge engineering in soil erosion

Bernard Allen Engel, Purdue University

Abstract

Concepts for integrating knowledge from several sources were proposed and implemented in a program called MKSMART (Multiple Knowledge Source Management And Reasoning Tool). MKSMART uses artificial intelligence techniques to allow the construction of blackboard-like applications. It is also useful for coupling numeric and symbolic reasoning; integrating software written in different forms; and as a problem formulation tool. The best knowledge representation and reasoning scheme can be used to solve components of problems. Conflict resolution strategies were developed and implemented within MKSMART to resolve conflicts between knowledge sources, which contain the domain knowledge needed to solve problems. MKSMART is modular allowing knowledge sources to be easily added, removed, or modified. A soil erosion model was developed using MKSMART and artificial intelligence programming techniques. The model demonstrated the usefulness of MKSMART and that of artificial intelligence in modeling soil erosion. MKSMART successfully combined knowledge in a number of forms from several sources to achieve the intended research objectives. The soil erosion model was based upon relationships developed by the USDA-ARS Water Erosion Prediction Project (WEPP). Knowledge sources in addition to those required to implement the erosion model were developed to provide model input and to interpret the model results. A graphical interface provides assistance in using the model, accepts inputs, and displays model results.

Degree

Ph.D.

Advisors

Beasley, Purdue University.

Subject Area

Agricultural engineering|Artificial intelligence

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS