Parameter induction underdamping steady-state matching stochastic resonance weak feature enhancement method

Abstract

A parameter induction underdamping steady-state matching stochastic resonance weak feature enhancement method comprises: employing the Hilbert transform to perform demodulation of obtained vibration signals, releasing fault feature frequency to a low-frequency area, and obtaining a corresponding envelope; inputting a parameter induction underdamping steady-state matching stochastic resonance system, taking a weighting signal-to-noise ratio of system resonant response as a target function of the quantum-inspired genetic algorithm, optimizing a system parameter, a damping ratio and a scale factor, triggering a particle movement mode and resonance cooperation between a transition rate and system input, and allowing the fault feature frequency to be located in a narrow transmission band of a stochastic resonance nonlinear band-pass filter; and setting a parameter induction underdamping steady-state matching stochastic resonance system according to an optimal parameter pair, inputting the envelope into a set stochastic resonance system, calculating and obtaining a resonance response of the system and performing analysis, and realizing enhancement and extraction of a mechanical fault feature frequency. The parameter induction underdamping steady-state matching stochastic resonance weak feature enhancement method improves weak feature enhancement and extraction capabilities of stochastic resonance.

Description

technical field [0001] The invention belongs to the technical field of mechanical fault diagnosis, and in particular relates to a parameter-induced underdamped steady-state matching stochastic resonance weak feature enhancement method. Background technique [0002] With the complex structure of mechanical equipment and the harsh working environment, it is difficult to enhance and extract the weak features of early faults, and it is difficult to realize the condition monitoring and timely maintenance of equipment. As a powerful tool for fault diagnosis of mechanical equipment, signal processing can realize fault diagnosis and maintenance of equipment by revealing early fault symptom information hidden in vibration signals. As one of the noise-assisted signal processing methods, stochastic resonance can use nonlinear systems to capture noise energy enhancement and extract weak features of early mechanical failures. This concept of turning waste into treasure makes stochastic r...

AI Technical Summary

Problems solved by technology

[0004] However, the single steady-state overdamped stochastic resonance induced by traditional parameters has the following disadvantages: (1) Different working conditions and fault types will inevitably lead to different vibration signals obtained at different times and sensor locations, so it is difficult to achieve a single steady-state stochastic resonance with multiple Effective steady-state matching between variable vibration signals; (2) The noise response of overdamped stochastic resonance has a Lorentzian distribution form, that is, the noise energy gathers in the low-frequency region, making it difficult to suppress multi-scale noise, which needs to rely on high-pass filters preprocessing

If the filter parameters are artificially set improperly, it may cause system output divergence or false detection and false alarm of faults; (3) The scale factor determines the response frequency band characteristics of the stochastic resonance system, and the fixed scale factor is likely to make it difficult for the stochastic resonance system to capture the most noise energy Realize the enhancement of fault characteristics, even the excessively wide passband causes a large amount of background noise to remain in the resonance response of the system, which interferes with the extraction and identification of fault characteristics

These shortcomings have limited the ability of traditional parameter-induced single steady-state overdamped stochastic resonance to enhance and extract early fault weak features

Images