The Application Of Wavelet Threshold De-noising In The Aluminum Reduction Cell Electrolyte Temperature Data

In electrolysised aluminum production, computer records, manual measurement and analysis test may create large amounts of data. The internal relationship in these data that can be excavated will become valuable knowledge, which will provide the sicentific style production and management of electrolysised aluminum production with great supports and help. However,these data may contain unstable signals which need to be processed before date mining and date analysis.This paper systematic study the Haar wavelet, Db2, Db3, Db4 in Db series of wavelet and Syml, Sym2, Sym3 in Sym series of wavelet algorithm and application. And the Wavelet algorithm is proved and applied to data analysis of electrolysised aluminum. Finally, it will be analysised by K-means Clustering Algorithm.1. Design and complete data management capalibilites,including:Data accessing, basic data maintenance and data pre-processing2. The traditional Haar wavelet, Db series of wavelet and wavelet denoising algorithm of Sym series are comparatively researched and analysised.3. Combine the thinkings of wavelet denoising and wavelet packet denoising and improved wavelet denoising algorithm, and applied it to the electrolysised aluminum data denoising.4. Chart the original data, preprocessing data, the traditional wavelet after denoising and inproved wavelet denoising.5. Apply K_means clustering algorithm to aluminum-electrolysised production data, and analyisis the mount and level of aluminum and electrolyte level and study the relationship between them.