Wavelet Associate Fractal Analysis Vibration Signal Of Roller Bearing

With the modern industry and the science and technology swift and violent development, automatic degree is higher and higher, this set the new request to the equipment maintenance. Structure of the mechanical equipment is complex day after day, because once the equipment breaks down, often cause serious consequence, not only block production, further more can form enormous harm to the operator. State estimation of mechanical equipment is an absolutely necessarily link of the mechanical equipmentt fault diagnosis technique, it is refers according to the historical information of equipment movement parameter determined the running status of the equipment at present, forecast future run tendency of the equipment, and will carry on the estimate to the surviving life of equipment. Carries on the forecast to state of mechanical equipment, is precondition to carry on foreknow maintenance.The rolling bearing is one of widespread application parts in the modern machine. As one kind of general mechanical apply the most wide range part for each kind of rotating machine, its running status is whether or not in gear often directly touch performance of the integral machine (includes precision, reliability and life and so on). This article has carried on the analysis to the bearing equipment and the complicacy of its breakdown with viewpoint of systems science and Fractal theory.Due to improper arrangement, poor lubrication, wet and foreign bodies incursion, corrosion and overload and so on possibly be able to cause the rolling bearing premature failure in the generating process. Even in the installation, lubrication,useing and maintenance all under normal circumstances, after operating for some time, bearing may also appear fatigue desquamate and attrition wear,and not normally wear and spalling. The following are the main cause of the malfunction:Fatigue desquamate, wear, plastic deformation, corrosion, fracture, agglutination, Cage damage. In the process of rolling bearings operations, vibration come into being due to impact of rolling element with the inner or outer body. By this the vibration frequency was natural frequency of the various parts of the bearing. Bearing vibration can be caused due to external vibration source, it may also be due to the bearing structure characteristics and defect. Bearing vibration can be caused due to external vibration source, it may also be due to the bearing structure characteristics and defect. In principle, it can be divided into flexible with the vibration of bearings and bearing rolling surface conditions (Bellows. injuries, etc.) of the two types of vibration. The former has nothing to do with the unusual state of the bearing, on behalf of the vibration system transmission characteristics to a certain extent. The latter has reflected the damaged state of the bearing. In addition, also perhaps there are fluid motive force vibration and noise that are produced because of lubricant in the course of operation of the bearing. Main failure mode in rolling bearing is it stand up one is corrosive pitting, desquamate and abrasion, etc. to include for surface damage to summarize, another kind is a kind of abrasion failure. Because abrasion failure will not cause the bearing to destroy at once, its extent of injury is smaller than the surface damaging type trouble far, so usually what people paid attention to most is still surface damaging type trouble.The time course of the signal can be used for describing the amplitude of signal is up to the change of time; The frequency spectrum of signal can express information of the amplitude of signal varies with frequency. The ideal tool which analyses smooth signal is Fourier transform. To the no-stationary signal, because Fourier transform can't describe the partial frequency characteristic of signal, is no longer an effective analysis tool. Wavelet is exactly a strong tool which analyses the no-stationary signal. In theory, the place can use Fourier transform than also can use wavelet analysis besides asking the question of solving of differential equation, and can obtain better result. Wavelet has enormous advantages in signal processing, lie in the wavelet transform can achieve multiresolution describe of signal, meanwahile, it has abundant wavelet basis that t can meet the signal with different characteristics. Utilize the wavelet to sublimate the vibration signal of the bearing, has got rid of the noise signal and created conditions for analysis of the trouble signal of the bearing.It is very irregular and unsuitable to describe with classical geometry to assign to the target described of fractal. Fractal meaning is"broken, cracked", also have"irregular meaning".Generally speaking, fractal mean have limitlessly fine, very irregular, infinite similar structure and some negation integer point set. Say roughly, it is geometry of an irregular form, but this kind of scrambling (coarse) has hierarchy, namely can be observed under the different levels (yardstick). In fact, irregular geometry abstract can clench the teeth natural world accurate than smooth rule whole geometry in classical geometry. It is that some "complicated points"gather on some simple space to fractal, this kind of set has some special nature, particularly, fractal has two pieces of typical geometry nature: self similarity and free scaling property. It uses an important parameter that the ration portrays the Chaos and attracts son's" strange degree" to divide the shape dimension (fractal dimension), widely applied to portray the digital characteristic of the non-linear systematic behavior, it is one of the cores of dividing the shape theory. In a great deal of points of shape dimension, the related dimension , to attracting the son's disparity reflection to be most sensitive, we pass the related dimension while analysing different working state of the rolling bearing, to study the relation between related dimension and bearing working state. Examine because of us vibration that have signal dispersed time series, as to how try to get related dimension, time of array, Grassberge and Procaccia were according to imbedding the theory and rebuilding the space thought of the looks in 1983, proposed calculating the algorithm of the related dimension from the time array data directly, namely G-P algorithm is also called " imbedding the space law ". We know , the same signal imbeds the dimension differently, its related dimension is different, it is confused not to undergo the related dimension of the signal dealt with in noise reduction and imbed the relation between the dimension , and related dimension its imbed dimension have certain laws after the noise reduction, because the existence of the noise influences the result of calculation of the related dimension directly, the related dimension with noise signal can't describe the systematic characteristic accurately. So, must deal with noise reduction to the signal. We can use related dimension as the characteristic amount to signify the systematic state. Divide the introduction of the shape theory , has offered the new effectual way for settlement of a lot of problems in the mechanical engineering field, have promoted the mechanical deep development in several respects of science and technology.Had on the basis of the understanding of basic conception and applying to project of fractal, give out fractal algorithm which adapt to signal processing of mechanical signal, on the basis of fractal analyse to the analog signal, have fractal to analyse, finding the signal counts the operation state that can describe the bearing clearly about the vibration signal of the bearing, this kind of method has succinct , quantitative number value index.Fractal can be used for analysing other mechanical vibration signals. Have combined the wavelet with fractal to the real bearing vibration signal and carried on research, calculate out the related dimension after picking out the noise signal use wavelet and constructing again the looks space , to judge the bearing whether there is a conclusion of the trouble. Utilizing wavelet and fractal can carry on characteristic analysis to the trouble signal of bearing, this conclusion is certain directive significance to diagnose the trouble of the bearing. Method this in the same way, it have very important meaning too to other mechanical failure diagnosis.