Application of Fuzzy Neural Network to Predict Irrigation Water Distribution: An Example with Cho Main Canal system

• Main Authors: LIN,SHENG-HAO, 林聖晧
• Other Authors: CHEN,YI-CHUNG
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/t2r857

碩士 === 國立雲林科技大學 === 工業工程與管理系 === 107 === Global warming and the resulting climate change are increasing the frequency of extreme weather in Taiwan, and water restrictions have been implemented throughout Taiwan in recent years. To minimize the losses caused by drought, irrigation associations all over Taiwan have been actively studying the means of optimally distributing and dispatching existing water resources. In the past, water distribution and dispatch relied on the experience and judgment of the managing personnel, which can be subjective and result in uneven distribution. This study therefore proposed a method based on a fuzzy neural network to optimize water distribution. We first collected data on main and branch distribution, type of crop, and the size of irrigated land and converted the data to unify the formats. The coded main and branch dataset was then used for water distribution. The proposed approach can calculate how much water is distributed to any irrigation waterway in an irrigation system comprising one or many branches. We performed a simulation experiment using an actual irrigation branch dataset to demonstrate the reliability and stability of the proposed method in the prediction of irrigation water distribution. Keywords: Water Distribution, Fuzzy Neural Network, Artificial Intelligence