Article ID: | iaor1993582 |
Country: | United States |
Volume: | 117 |
Issue: | 4 |
Start Page Number: | 418 |
End Page Number: | 434 |
Publication Date: | Jul 1991 |
Journal: | Journal of Transportation Engineering (ASCE) |
Authors: | Roschke Paul N. |
Keywords: | artificial intelligence: expert systems |
A prototype knowledge-based expert system for assisting engineers in selecting, locating, and designing construction and retrofit projects for traffic barriers, such as roadside barriers, median barriers, bridge rails, and crash cushions is described. The system uses design principles of traffic barriers, information provided through interviews with experts, and information on a particular site on generate feasible candidate designs. These designs are processed by an evaluation system that includes finite-element analysis, a weighting mechanism, and a cost-benefit model. Color graphics and explanation module illustrate the design. An integrated rule- and frame-based expert-system shell facilitates implementation of the knowledge-based system on a microcomputer with the help of external routine links, data-base access, and a graphical interface. Visual knowledge, which is preserved through scanned images, is replete with educational benefits to highway engineers. An example bridge-rail-retrofit procedure demonstrates the approach to development of the knowledge-based system.