Gas pressure regulating station state monitoring method based on optimization feature selection

Abstract

The invention discloses a gas pressure regulating station state monitoring method based on optimization feature selection, and the method comprises the steps: firstly converting a distance between a comparison matrix and a reference matrix into a probability through geometric probability distribution, and solving a problem that data initialization translation is needed before gray correlation analysis; then, a self-adaptive value taking method of a resolution coefficient is proposed based on a cuckoo algorithm, finally, a weight coefficient is constructed in combination with redundancy between features and relevancy between the features and classes, the grey relational degree used for a non-sequential system is improved correspondingly, the algorithm performance is verified through multiple different types of data sets, and the algorithm performance is verified. The performance of the feature selection method is compared by adopting two performance indexes, namely the maximum classification accuracy and the number of selected features corresponding to the maximum classification accuracy, the universality and superiority of the algorithm are verified, finally, the method is applied to state monitoring of the gas pressure regulating station, the feasibility and effectiveness of the method are verified, and dimension reduction of multi-dimensional information is achieved.

Description

technical field [0001] The invention belongs to the technical field of state monitoring of gas pressure regulating stations, and in particular relates to a state monitoring method for gas pressure regulating stations based on optimization feature selection. Background technique [0002] With the development of the Industrial Internet, the safety and economy of industrial production control has become the focus of attention. As one of the industrial control equipment, the gas pressure regulator is an important part of gas transmission and distribution. It can automatically adjust and stabilize the pressure of the pipeline network, and play a very critical role in the normal and safe operation of the entire gas pipeline network. The research on working condition monitoring of pressure regulating station has important practical significance. [0003] The condition monitoring of the gas pressure regulating station is usually through the analysis of its outlet pressure signal. ...

AI Technical Summary

Problems solved by technology

The premise of these methods all need to initialize the data, but in the working condition identification of the gas pressure regulating station, some features can represent different working conditions when the difference is small. If the data is shifted, although the data The distribution shape has not changed, but the working conditions it represents may change; if the absolute distance is used to process all the features, it is not completely suitable for some features with large differences

Therefore, the existing gray relational analysis method is not fully applicable to deal with the characteristics of the gas pressure regulating station

Images