Please use this identifier to cite or link to this item: http://cmuir.cmu.ac.th/jspui/handle/6653943832/69249
Title: การพยากรณ์ระดับน้ำในสถานีวัดน้ำโดยใช้เทคนิคโครงข่าย ประสาทเทียม
Other Titles: Development of Water Level Prediction System in Water Level Measurement Stations Using Neural Network
Authors: อาจารย์ ดร.ภาสกร แช่มประเสริฐ
บุญนาท สุวรรณศรี
Issue Date: Apr-2016
Publisher: เชียงใหม่ : บัณฑิตวิทยาลัย มหาวิทยาลัยเชียงใหม่
Abstract: This independent study is to study and apply the artificial neural network model to predict the water level in the river for flash flood preparedness. The experimenting with the water level in the past with different periods (i.e., 10 minutes, in average of one hour, and at one hour) is applied to predict water level in the future. Moreover, the water level data from two monitoring stations is applied to evaluate the predicted water level and compared with the data from only one station. The simulation results show that the predicted water level at 10 minutes is accurate as the average error is lower than the set criteria of 0.05. It is also found that the 3 major factors affecting the water level prediction using the artificial neural network, which are the numbers of data, the numbers of hiding levels and the time of prediction. It also reveals that the higher the number of data, the more accurate or precise it could be; however, it takes a long time. The more numbers of hiding levels, the more time it took the artificial neural network for the measurement. The lesser the prediction time, the more precision the prediction is. However, the factors of the artificial neural network (e.g., the number of nodes, hiding levels, the number of input data) is also required to be carefully designed.
URI: http://cmuir.cmu.ac.th/jspui/handle/6653943832/69249
Appears in Collections:ENG: Independent Study (IS)

Files in This Item:
File Description SizeFormat 
Full.pdf4.28 MBAdobe PDFView/Open    Request a copy


Items in CMUIR are protected by copyright, with all rights reserved, unless otherwise indicated.