| Summary: | 碩士 === 國立屏東科技大學 === 水土保持系 === 94 === Traditional hydrology is a science of probability and mathematical statistics to be analyzed and estimated on water resources management, disaster prevention and flood prevention design. But the records of hydrologic have an attribute of high uncertainty. The estimation could be affected by record range, unit scale, position of rainfall observation station, even human negligence. Thus, the natural hydrologic phenomenon may include hypotheses and misunderstanding using Binary Value logic method. The topic of this study is to surmount the existing obstacles by simulating AI analysis method on hydrologic estimation and decision.
The research of the study is based on the Fuzzy Theory. Data were collected from PingTung Area around Don-Kung watershed with a viewpoint of scaling, thru the 38 rainfall observation stations. The rainfall types included hourly, monthly and yearly records. The important conclusions of the study are stated as follows：(1)The rainfall observation stations can be reduced to 12 from 38 on the hydrologic characteristics of monthly average rainfall at Don-Kung watershed. (2)Spatial distribution using Kriging method at Don-Kung watershed shows that the country yearly average rainfall are 1,939mm, 1,991mm, and 2,031mm by all observation stations(No.38), overall historical yearly records(No.16), and the stations to be chosen(No.25). The study of rainfall events which were over 100 mm summed 1,138, is with the characteristics of long duration and platy kurtosis. The rainfall observation station C1R12 has the nearest figures with the average rainfall characteristics in PingTung.
|
|---|
