Time Series Mining Algorithms In The Slot Status Of Judgment And Research

In the existing production process of aluminum electrolysis, a large number of electrolytic cell data, which reflected the slot working status in the sampling instant and provided users with services through graphs, statements and other forms at the host computer monitoring system, were collected by the control system. However, these massive data only provide users with the slot working status in various periods. Users must rely on their own professional knowledge and rich experiences to carry out preliminary qualitative analysis to obtain information out of graphs, statements, etc. A lot of useful information were hidden under these massive data. It is important to analyze these data effectively to obtain the status and developing trends of single slot and a series of slot, it is also important to classify all the cell and realize different management according to different classification in the producing process. At present, there is still a lack of work in this area.This paper presents series of data based on sliding window analysis algorithm, to solve the analysis and forecast problem of slot status in production of aluminum electrolysis. The core of this analysis algorithm is the construction of the special connective branch based on the distance of points, and judge the status of the current trough using the sliding window of time. This paper gives detail discussions of basic concepts of the algorithm, and gives the corresponding branch connectivity mining algorithm. Through the connective branch, the author carry out cluster analysis of data, obtain the classification of the status of series of slot, and bring forward time sliding window approach to analyze the status and developing trends of single slot.This paper gives cluster analysis and forecast analysis to the massive data arising from the producing process of aluminum electrolysis, then obtain outcome of the status of series of slot and single slot, thus guiding the production. This is of positive significance, and also the problem urgent to solve in the actual production. Through the experimental analysis of actual production data, there are strong guiding significance to actual production.