The Study On Dynamic Characteristics And Diagnosis Methods Of Gear Backlash Fault

In the modern manufacturing industry, rotating machinery damage always has a relatively dramatic impact signals such as broken teeth, tooth cracks, bearing failure. Therefore, the impact signal may be judged by the condition of the equipment and the specific fault, while the impact of the transient signal is often drowned in noise and other signals, so the extracting of the impact of the transient signal is signal processing of a problem. Using traditional signal processing method is difficult to effectively extract the shock signal, the existing extraction method of transient signal most need for human intervention, if can find a way to automatically extract the transient signal is a great significance.This paper, around the ear auditory features automatic extraction of impact signal as the center was studied, the main research content is as follows:(1)Establish the human ear auditory model. First of all, according to the biological understanding of the human ear peripheral model, selects the Gammatone filter group as a peripheral auditory models. Using this model,different frequency signals was separated which conducive to the further analysis (such as shown in the output of the time-frequency we can see amplitude modulation signal; preliminary can make the impact of the weak signal.). Then, this paper adopts Meddis model to simulate the auditory system of inner hair cells. The main function of Meddis model is half-wave rectifier, nonlinear compression and adaptive adjustment, can reduce the amount of data the next step for signal processing.(2) The transient signal extraction. The characteristics of impact signal is transient, characterized by two aspects:the instantaneous frequency, instantaneous amplitude. In this paper, using the envelope method to extract the instantaneous amplitude value, can be more prominent instantaneous amplitude value,and reduce the amount of calculation. The signal amplitude of each channel two envelope (amplitude modulation, frequency modulation and sine signal can be directly eliminated by two envelope), amplitude of the second envelope is concluded the starting point of impact signal;The impact of envelope signal end points can be concluded by Minimum operation, so we can put the signal of impact signal is extracted. (3) The simulation and experiment. Set the secondary gear reducer test bench, can achieve a variety of gear fault simulation. This article is mainly used for simulating pitting and broken teeth gear fault. Movement in this paper, the method successfully extract the transient signal from the simulation signal and test fault signal of the ingredients.(4) The transient signal spectrum estimation. The transient signal amplitude spectrum contains the natural frequency of defective parts, the transient signal spectrum estimation can improve the estimation precision of the natural frequencies of the fault parts, which parts failure are obtained.