A knowledge-based system for highway cutting slope design

Bedford, John Michael (1991) A knowledge-based system for highway cutting slope design. University of Southampton, Doctoral Thesis.

Abstract

The thesis describes the design of a prototype knowledge-based system that successfully demonstrates the relevance of this technology to the design of new highway cutting slopes and remedial measures in high plasticity overconsolidated clays. Highway cutting slope design is a complex geotechnical design application that requires both numeric and symbolic approaches to achieve design solutions. Knowledge-based system technologies address the symbolic requirement. The specific cutting design strategies investigated for the prototype system are a preliminary design strategy, a design strategy for first time slides (Chandler and Skempton method) and a design strategy based on laboratory derived strength parameters and a suitable method of analysis (i.e. the simplified Bishop method). In addition, a strategy is presented for generating the design of remedial works. The prototype system incorporates both rule and script knowledge representations for different types of knowledge found in geotechnical design. Within an overall data-driven reasoning strategy there is a requirement for both backward and forward reasoning in cutting slope design. As input data from which an engineer makes design judgements can be fuzzy, an assumption based approach for modelling beliefs is adopted. Belief intervals are represented using system time. The IKBS approach is shown to extend the role of the computer in geotechnical design and allow the integration of existing numerical procedures. This has potential uses in both computer based training and as a decision aid for the design engineer.

This record has no associated files available for download.

Contributors

Download statistics