Draught fan failure detection apparatus and method

Abstract

The invention discloses a draught fan failure detection apparatus and method. The draught fan failure detection apparatus structurally comprises a position adjusting unit and a vibration acquisition frame arranged on the position adjusting unit. A displacement sensor connected to a processing host is arranged on the vibration acquisition frame. The draught fan failure detection method with the draught fan failure detection apparatus includes the steps of A adjusting the position of the displacement sensor; B, simulating draught fan failure and collecting a failure signal; C, obtaining a time series of the failure signal; D, converting the time series into a symbol sequence; E, calculating information entropy of the symbol sequence; F, extracting a feature vector; G, performing a first model training; H, performing a second model training; I, obtaining a final training model; J, formally running a draught fan and comparing the signal with the final training model to obtain failure information. According to the draught fan failure detection apparatus and method, a vibration acquisition device is arranged reasonably and symbolization processing is performed on the acquired vibration signal to obtain a comparative database and reduce the impact of noise, so that the computation speed is greatly improved.

Description

technical field [0001] The invention relates to the field of early warning of fan failures, in particular to a device and method for detecting fan failures. Background technique [0002] In a power plant, the operating status of the wind turbine is directly related to the safe and economical operation of the power plant, and the reliability, safety and economy of the wind turbine depend on its efficient operation, real-time status tracking evaluation, accurate fault diagnosis and maintenance, so research Fault diagnosis of fan is of great significance. Common fan mechanical faults include rotor imbalance, rotor misalignment, bearing looseness, dynamic and static friction, etc. Although there are many fault diagnosis methods, they are basically divided into three steps: acquisition of diagnostic information; fault feature extraction; state identification and troubleshooting. [0003] Traditional signal feature extraction methods presuppose the stationarity of the signal, an...

AI Technical Summary

Problems solved by technology

[0003] The traditional signal feature extraction method is based on the stationarity of the signal, and cannot effectively analyze and process the non-stationary signal.

Due to the nonlinearity of the driving force, damping force and elastic force in the operation of the fan and the nonlinearity of the mechanical system, the detected vibration signal is a non-stationary signal, and the traditional signal feature extraction method has great limitations.

Images