三維廠區污染源之數值模擬

Abstract

目前廠區污染的偵測大多是擺設大量的監測點，並認為最高監測數據之監測點即為污染源位置，對於以有限監測點監控廠區，依據有限數據預測污染源位置，卻很少有論文研究。本研究的主要目的在於研究依據有限監測點下，有效預測污染源位置的方法。 本文的數值方法是根據擴散模式，參照統御方程式解析解的形式，假設所需近似曲線形式，以此近似曲線求出監測點連線上最大值位置，最後利用各條近似曲線最大值位置的法線方程式找出洩漏源位置。再以處理二維問題的經驗建構符合三維問題的監測網路。 不同於其他文獻大量使用監測點，此法利用有限個監測點所得的已知濃度資料，能有效地反推出監測範圍內洩漏發生的位置。

The aim of the study is to predict the location of pollution source with 13 sensors and establish a mathematical model. According to the form of the diffusion governing equation, the mathematical model is to find a fitting approximate curve equation. At first, a physical model simulates the pollution situation computed by CFD software. The computed concentration data on specifically arranged locations are treated as the known detective data. Then, analyze the data by the mathematical model to obtain a fitting curve equation. Consequently, the intersection point of the normal equations of the maximum value on each fitting curve can be regarded as the location of pollution. Finally, construct the network of prediction used in 3-D problems form the experiences dealing with 2-D problems.