Industrial equipment fault detection method with time series state variable

Abstract

The invention discloses an industrial equipment fault detection method with a time sequence state variable, and belongs to the field of industrial equipment fault detection. The problem that fault detection of an existing industrial system depends on fault data to establish a detection model, fault data is difficult to obtain, and the method is difficult to apply is solved. Comprising the steps of collecting time sequence state data as training data; calculating a mean value and a standard deviation of m-dimensional characteristic variables of the training data, and standardizing the training data; segmenting the standardized time sequence data by using a sliding window to obtain sequence data P, and training the LSTM self-encoding network; obtaining an error sequence of training data based on the trained LSTM self-encoding network; analyzing the error sequence based on an extreme value theory method to obtain a fault threshold value; and the fault threshold value is adjusted based on the sequence data of the to-be-detected equipment to obtain an adjusted threshold value, so that fault detection of the industrial equipment is realized. The method is used for fault detection of industrial equipment with variables having time sequence characteristics.

Description

technical field [0001] The invention relates to an industrial equipment fault detection method with time series state variables, belonging to the field of industrial equipment fault detection. Background technique [0002] Fault detection has a wide range of applications and an important position in the industrial field. Industrial systems and equipment usually have extremely complex components, and their failure will bring huge harm and loss, so it is of great significance to monitor and detect faults. [0003] At present, it is extremely difficult to detect the fault of industrial equipment. On the one hand, the industrial system has a huge composition and complex structure, and the parameter dimension of the monitoring data is high; and the inherent characteristics of the system are contained in the complex parameter relationship, so traditional feature extraction is adopted. It is difficult for the method to extract features that can reflect its intrinsic performance; o...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that the fault detection of existing industrial systems relies on fault data to establish a detection model, and the difficulty in obtaining fault data makes the method difficult to apply, the present invention provides a fault detection method for industrial equipment with time series state variables

Images