Research On Parameters Deterioration Of Auxiliary Equipment In Power Plants

The change of auxiliary equipment from normal operation to fault state is not mutated, but gradual and has rules to follow. During the deterioration of faculties of the auxiliary equipment, it is meaningful to find its symptoms in early stage. However, nowadays, the traditional equipment alarming system works only when those parameters reach alarming value, but this kind of alarming method cannot detect symptoms which indicate the slow change of faculties of the equipment with effect. Auxiliary equipment in power plants product quantities of data every day, such as temperature, vibration, electric current, pressure and so on. These dates also reflect the operation state of the equipment. It is a pity that a number of historical data during the operation of auxiliary equipment stored in the real-time and historical database has not been applied well. Meanwhile, as a result of the effect that some factors which are not considered or are not able to be considered will influent the change of operation parameters of auxiliary equipment, it is impossible for us to use precise functions to describe relationship among variable quantity of auxiliary equipment. According to this kind of condition, this thesis will research monitor methods of deterioration of auxiliary equipment at early stage from the point of statistics independent of mathematic model, based on those historical data of some important parameters during its operation. Specifically, they are researched from three aspects, namely univariate, multi-variable and spatial distance. Univariate is the method which researches the trend of equipment deterioration from the aspects of concentration tendency, dispersion tendency and the shape of data distribution. Multi-variable method researches from the points of correlation, principle component analysis and factor analysis. Analysis of deterioration based on the spatial distance is the method that makes use of the principle that equipment parameters aggregate in a certain spatial location when the equipment runs normally.