Reciprocating mechanical equipment fault detection method based on real-time vibration data

Abstract

The invention discloses a reciprocating mechanical equipment fault detection method based on vibration data, and relates to the technical field of diesel engine operation state monitoring. The method comprises the following steps: firstly, aiming at target equipment, optimizing a measuring point channel which is most sensitive to an operation state through working condition sensitivity analysis; in the actual operation process of the target equipment, collecting a vibration signal of the optimal measuring point channel in real time; performing Fourier transform on each new data segment to obtain a Fourier coefficient vector; continuously arranging the Fourier coefficient vectors of different data segments in the time direction to obtain 'time-frequency-amplitude' three-dimensional joint distribution of the vibration data of the channel, and constructing 'time-frequency-amplitude increment' three-dimensional joint distribution and enhanced 'time-frequency-amplitude increment' three-dimensional joint distribution; and finally, comparing with operation condition data, and detecting fault characteristics. According to the invention, information extraction can be effectively carried out on reciprocating mechanical equipment in real time, and faults can be found and positioned in time.

Description

technical field [0001] The invention relates to the technical field of diesel engine operation state monitoring, in particular to a method for detecting faults of reciprocating mechanical equipment based on vibration data. Background technique [0002] Typical reciprocating mechanical equipment such as diesel engines and reciprocating pumps have the characteristics of complex structure, precise mechanism, and high-speed movement. Reciprocating mechanical equipment is usually used as a power source or driver and is an important equipment in the system. However, in actual engineering, reciprocating mechanical equipment often has poor installation environment, long continuous working hours, high reliability requirements, and difficult repairs. Fault detection is essential for reciprocating mechanical equipment. Good detection methods can effectively avoid major faults and cause serious economic losses. Generally speaking, the requirements for fault detection in engineering are...

AI Technical Summary

Problems solved by technology

This method has obvious deficiencies such as poor failure, strong randomness, and incomplete testing, resulting in large errors in test results and untimely fault detection.

[0004] In addition, typical reciprocating mechanical equipment such as diesel engines and reciprocating pumps often require continuous signal monitoring and analysis. However, traditional time-domain statistical methods are sensitive to operating conditions, and can only preliminarily judge whether there is an abnormality, and cannot quickly locate faults; Theoretically, the spectrum analysis method is only suitable for constant speed and load conditions, and the Fourier spectrum does not contain time information, so it cannot be compared before and after, which is easy to cause false alarms; the time-frequency analysis method can express the joint information of time-frequency-amplitude , but involves two-dimensional matrix calculations, poor real-time performance, and the bottleneck of time-frequency resolution that cannot be optimal at the same time. There is no perfect time-frequency analysis method for engineering

Images