The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks
碩士 === 國立雲林科技大學 === 工業工程與管理系 === 103 === According to the literature, each Taiwanese receives only one-eighth of average rainfall per year in global. Esspaically the situation for Shihmen Reservoir which many compatriots depend on, that makes administration of water resource become more difficult an...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2015
|
Online Access: | http://ndltd.ncl.edu.tw/handle/88713547262607534067 |
id |
ndltd-TW-103YUNT0031032 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-103YUNT00310322016-07-02T04:21:21Z http://ndltd.ncl.edu.tw/handle/88713547262607534067 The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks 應用類神經網路建構石門水庫入流量與放流量之預報模式 Sin-Hong Yu 余欣虹 碩士 國立雲林科技大學 工業工程與管理系 103 According to the literature, each Taiwanese receives only one-eighth of average rainfall per year in global. Esspaically the situation for Shihmen Reservoir which many compatriots depend on, that makes administration of water resource become more difficult and important. Due to the flow of reservoir is one of the important foundation data for decision management, the purpose of this study is to establish the prediction mode of flow to aid decision management. For administration, this study looks forward to reduce extra resources wasting of scheduling or purchasing water, and reduce disaster risk caused by misjudgment decision. For environment, this study expects stable flow be able to avoid high turbidity water and landslide caused by torrent. Shihmen Reservoir is research target in this study. The predictive variables of high explanatory power will be found though all possible regression procedure, then combine with Back Propagation Neural Networks and Time Delay Neural Networks respectively to construct forecasting modes. Then, to raise the accuracy and reliability of the prediction mode by combining with Time Delay Neural Networks and Nonlinear Auto-Regressive Moving Average control theory. At Final, evaluating accuracy of mode by error analysis and coefficient of efficiency. As a result, NARMA-L2 is the best mode. Back Propagation Neural Networks is better than other research the same method due to different input variables. Therefore, the study provides different methods and input variables for future research reference. Tung-Hsu Hou 侯東旭 2015 學位論文 ; thesis 95 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 國立雲林科技大學 === 工業工程與管理系 === 103 === According to the literature, each Taiwanese receives only one-eighth of average rainfall per year in global. Esspaically the situation for Shihmen Reservoir which many compatriots depend on, that makes administration of water resource become more difficult and important. Due to the flow of reservoir is one of the important foundation data for decision management, the purpose of this study is to establish the prediction mode of flow to aid decision management. For administration, this study looks forward to reduce extra resources wasting of scheduling or purchasing water, and reduce disaster risk caused by misjudgment decision. For environment, this study expects stable flow be able to avoid high turbidity water and landslide caused by torrent.
Shihmen Reservoir is research target in this study. The predictive variables of high explanatory power will be found though all possible regression procedure, then combine with Back Propagation Neural Networks and Time Delay Neural Networks respectively to construct forecasting modes. Then, to raise the accuracy and reliability of the prediction mode by combining with Time Delay Neural Networks and Nonlinear Auto-Regressive Moving Average control theory. At Final, evaluating accuracy of mode by error analysis and coefficient of efficiency. As a result, NARMA-L2 is the best mode. Back Propagation Neural Networks is better than other research the same method due to different input variables. Therefore, the study provides different methods and input variables for future research reference.
|
author2 |
Tung-Hsu Hou |
author_facet |
Tung-Hsu Hou Sin-Hong Yu 余欣虹 |
author |
Sin-Hong Yu 余欣虹 |
spellingShingle |
Sin-Hong Yu 余欣虹 The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
author_sort |
Sin-Hong Yu |
title |
The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
title_short |
The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
title_full |
The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
title_fullStr |
The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
title_full_unstemmed |
The Study for The Prediction of Inflows and Outflows of The Shihmen Reservoir by Using Artificial Neural Networks |
title_sort |
study for the prediction of inflows and outflows of the shihmen reservoir by using artificial neural networks |
publishDate |
2015 |
url |
http://ndltd.ncl.edu.tw/handle/88713547262607534067 |
work_keys_str_mv |
AT sinhongyu thestudyforthepredictionofinflowsandoutflowsoftheshihmenreservoirbyusingartificialneuralnetworks AT yúxīnhóng thestudyforthepredictionofinflowsandoutflowsoftheshihmenreservoirbyusingartificialneuralnetworks AT sinhongyu yīngyònglèishénjīngwǎnglùjiàngòushíménshuǐkùrùliúliàngyǔfàngliúliàngzhīyùbàomóshì AT yúxīnhóng yīngyònglèishénjīngwǎnglùjiàngòushíménshuǐkùrùliúliàngyǔfàngliúliàngzhīyùbàomóshì AT sinhongyu studyforthepredictionofinflowsandoutflowsoftheshihmenreservoirbyusingartificialneuralnetworks AT yúxīnhóng studyforthepredictionofinflowsandoutflowsoftheshihmenreservoirbyusingartificialneuralnetworks |
_version_ |
1718333474123808768 |