| Summary: | 碩士 === 國立中興大學 === 土木工程學系所 === 94 === Within a single river disaster prevention purpose, there could be many different construct methods can be used. For example, to deal with the disaster of river overflow, many engineering approaches could be taken, such as “heighten riverbanks” , “upriver slack”, “broaden river way ”, and so on. But which method is the best choice? Many factors should be considered, such as geologies, landforms, land-ownerships, the slopes of riverbed, contractors’ abilities, budgets, time-constraints, transportations, nearby-residents dispute, and so on. As a result, to making great decisions for river disaster prevention engineering methods required copious experience and knowledge.
After a long-term practice, a junior engineer could become a senior one with abundant experiences and knowledge. Then he or she knows how to choose suitable methods to solve problems. However the senior engineer will retire soon. How to keep the experience and knowledge of the senior engineers for young engineers becomes a very important issue.
This research is aimed to build up knowledge-database for river disaster prevention, through in-depth interview with experienced engineers (experts). We organized the obtained expert knowledge in the form of a matrix. The knowledge matrix not only stores artificial intelligent river disaster prevention knowledge, but also provides a mechanism for computers to retrieve knowledge and make good decisions. A web-based expert system has been established named “Artificial Intelligent River Disaster Prevention Expert System” (RDPES). The RDPES simulates the scenarios when a young engineer asks for advices from a senior engineer. The RDPES starts at prompting a series of questions like senior engineers. According the question are answered, the RDPES will propose engineering methods for river disaster prevention.
|
|---|
