The dynamic analysis theory of rolling bearing performance

The dynamic performance of the rolling element bearing analysis theory of 1. Rolling bearing performance of chaos analysis theory is an important method of modern nonlinear analysis - - - - - - - , rolling bearing friction torque and vibration with nonlinear dynamic characteristics, suit to use chaos theory is studied. Rolling bearing performance with chaos theory research, often in the concept of phase space reconstruction analysis of dynamic characteristics of nonlinear, time series with the mutual information method of time delay, with Cao method to calculate the one dimension time series, according to the time delay and embedding dimension of time series of Lyapunov index calculation, and the implementation of time series prediction. With chaos theory analysis of time series of attractor, calculation of time series saturation embedded dimension and estimate the correlation dimension, it is concluded that the median quantities of time series and to estimate the relationship between the correlation dimension, can comprehensively reflect dynamic characteristics of rolling bearing vibration, friction torque. However, the chaos theory in nonlinear dynamic analysis is not perfect. The analysis result is sensitive to the initial value of chaotic theory, there is a big difference between initial value can lead to different results, especially the discrete data is contained in the initial value, the result out of people. Before using chaos theory analysis, therefore, should be the data sequence preprocessing, reduce the influence of discrete data. Sound of data processing can effectively separate the data in the discrete data, and to deal with these data, to reduce or weaken the impact of these data on the data sequence. Based on this, this chapter will robust data processing and chaos theory to fusion, to vibration and friction torque performance for the specific case, analysis of nonlinear dynamic characteristics of rolling bearing performance. 2. Rolling bearing performance robustness of data processing in the process of rolling bearing performance experiment and test, manufacture, installation, lubrication and measuring instrument factors such as randomicity, uncertainty, will lead to the complexity of the rolling bearing performance data collection; At the same time, the experimental environment, the influence of temperature and other conditions, can also cause the deviation of data. Now most of the data analysis method the influence of the place. The method of the reasons: the first is the number who stand scatter some high to remove data as phase error play outliers, or ignore the high micro teaching, according to the measurement errors are not necessarily reactions may be true; Second is indeed has influence to the measurement, the data have different effect in different situations, and even decisive role to the change of performance, but should not be removed, should be careful consideration, the emotion processing. To this end, this chapter puts forward the improved Huber M estimation method to estimate the bearing vibration and friction performance, in order to reduce the influence of discrete data on the whole, and also contains the information of data. For example, for uniform distribution, binomial distribution, a common distribution such as normal distribution, the median and the mean value is consistent; For a large amount of data of the experimental data is similar to a certain distribution function, the median and mean value shall be the approximate therefore experimental data. If the difference in the median and mean that exist in the data variation or unreasonable data, can be used to robust estimation methods of Huber M data processing after the chaos analysis. The steps of this method are as follows: ( 1) According to the theory of statistical significance level of the data is given; ( 2) In sorting, the original data sequence from the most distant way in the median number of discrete data, replace with the adjacent data, get the new data sequence; ( 3) Analysis of median and mean value difference, differ of digits and average hours, a new data sequence is stable, then get a robust data; ( 4) Chaos analysis of robust data. 3. Rolling bearing dynamic performance analysis of the rolling bearing performance data fusion theory for steady, get the steady data, provides a reliable basis for data analysis. Rolling bearing performance characteristics of nonlinear, diversity and complexity, and chaos theory is one of the modern one of the important ways to solve nonlinear problems. Therefore, can use the chaos theory study, the properties of rolling bearing. But, chaotic system is sensitive to initial conditions, and the steady chemical experiment data processing can weaken the influence of the initial conditions, the organic combination of sound change theory and chaos theory, can effectively analyze the nonlinear dynamic characteristics of the rolling bearing problem. Steps are as follows: 1) To steady rolling bearing performance data processing, get the steady data; ( 2) Deal with rolling bearing performance robust data with mutual information method, obtain the time delay of phase space reconstruction; ( 3) With Cao method for rolling bearing performance robust data processing, get the phase space of the embedded dimension; ( 4) Based on rolling bearing performance steady data, time delay and embedded dimension, using chaos theory can get Lyapunov index; ( 5) Based on the rolling bearing performance steady data, time delay and embedded dimension, with the theory of chaotic attractor can be obtained; ( 6) Calculation of rolling bearing performance robust data saturation to one dimension, and estimate the correlation dimension; ( 7) It is concluded that rolling bearing performance data in the median number of physical space and estimate the correlation dimension of chaos.