嵌入引潮力與溫度的類神經-模糊潮汐推算模式

Abstract

本文主要探討引潮勢能與海溫參數對潮汐水位的影響，並應用類神經網路與適應性之模糊推算系統，架構潮汐水位推算模式。本文建構兩個水位推算模式，一為利用太陽及月亮與地球上對應測點之相對位置與距離等參數，輸入神經網路來建立以推算潮汐水位之模式為TGF-NN。另一為以海溫及引潮位勢進行類神經-模糊網路之架構是為TGFT-FN。 本文之TGF-NN模式與調和分析法、正交潮法及NAO.99b模式進行單一測站長時間推算水位之比較。本模式除了可應用於單點的潮位推算，亦應用其他測站的水位計算，說明本模式於工程應用上的實用性與精確性，推算之精度略優於NAO.99b模式。另外，嵌入海溫與引潮勢能之TGFT-FN模式，與引潮勢能之水位推算模式比較，TGFT-FN較TGF-NN模式於東部海域可修正方均根誤差約5公分，於西南海域則約可修正2公分；於西北海域兩模式之誤差接近。

The aim of this paper is to clarify how tide generating forces (TGF) and sea surface temperature (SST) affect the water level. This paper also uses back-propagation neural network (BPN) and adaptive network-based fuzzy inference system (ANFIS) to construct the tide predicting models. This study presents a neural network model, called TGF-NN, of simulating tides at multi points considering tide generating forces. The other model, called TGFT-FN, embedded tide generating forces and sea surface temperature is to calculate tides. A comparison on the root mean square and correlation coefficient of three year mixed tides at a single point computed with harmonic method, response-orthotide method, the NAO.99b model and the TGF-NN model was made to show the prediction accuracy of each method. This model is shown to be efficient as the harmonic method to estimate ocean tides at a single point. Extended application of this model to predicting tides at some points neighboring to an original interest point identifies accurately simulating multi-point tides as the NAO.99b numerical model. Furthermore, the difference of the model combined SST and TGF parameters and the TGF-NN model in eastern coast of Taiwan is about 5 cm, in southwestern coast is about 2 cm, and in northwestern coast is almost the same.