Long-term power-up equipment health state real-time anomaly detection method based on data driving

Abstract

The invention discloses a long-term power-up equipment health state real-time anomaly detection method based on data driving, and the method comprises the following steps: 1, carrying out the data processing of power-up equipment data collected by long-term power-up equipment, obtaining observation data, carrying out the conversion or enhancement processing of the observation data, and obtaining effective measurement data; step 2, carrying out dimension reduction processing on the effective measurement data in the step 1 to obtain data after dimension reduction processing, and selecting important variables from the data after dimension reduction processing by adopting a random forest model and correlation analysis; 3, according to the important variables obtained through screening in the step 2, life features are obtained through time domain feature extraction and frequency domain feature extraction; smoothing the service life characteristics to obtain the service life characteristics of the power-on equipment; 4, according to the service life characteristics of the power-up equipment in the step 3, adopting a fault trend prediction algorithm based on multivariable clustering and principal component analysis to obtain a fault factor; and 5, establishing a fault prediction model according to the service life characteristics of the power-up equipment calculated in the step 3. According to the invention, the prediction accuracy and the robustness of the model are improved.

Description

technical field [0001] The invention belongs to the technical field of health status detection of power-on equipment, and in particular relates to a data-driven real-time abnormality detection method for long-term power-on equipment health status. Background technique [0002] The test data of long-term power-on equipment is various data monitored by different sensors in real time during the operation of a certain equipment, and some parameters are directly or indirectly related to the operating status of the equipment. By analyzing long-term power-on data, a fault prediction and health management system (PHM) is built to evaluate the health status in real time, which can effectively guarantee the normal operation of the system where the equipment is located. [0003] The core and foundation of data-driven long-term power-on data health status monitoring lies in fault diagnosis and prediction technology. With the rapid development of test technology, especially the rapid im...

AI Technical Summary

Problems solved by technology

Machine learning and statistical analysis methods are the mainstream algorithms for data-driven long-term power-on equipment health monitoring. Existing health monitoring methods have problems of insufficient prediction accuracy and insufficient model robustness.

Images