Analysis Of Linearity For Seismic Noises Using DVV Algorithm

Nonlinear time series exist widely in daily life, signal nonlinearity is an important characteristic of the signal.in the recent twenty years.the world has been kept on studying the degree of nonlinearity in time series.Using nonlinear dynamics indexes can measure the sequence of nonlinearity, such as correlation dimension, Lyapunov exponent, three order autocorrelation function, NAR model, can reflect the potential kinetics in measured time series.We use time delay vector variance method to measure the nonlinearity of the measured sequence.use surrogate data to test the nonlinear components in time series, can obtain seismic noise nonlinearity more accuratly. If the measured DVV plot along the diagonal, then the measured sequence is linear; if the DW plot is off the diagonal.then the measured sequence is nonlinear.and the greater the degree of deviation from the diagonal, the greater the degree of nonlinearity.The delay vector variance method can be well used to measure the sequence of nonlinearity, but because we bring random phases in generating the surrogate data, the measurement of the same sequence may result in different DVV scatter plot.the stability of this algorithm is poor.To solve this problem we propose the improved delay vector variance method:1.Improved Time delay vector variance algorithm based on the null hypothesis:The sequence is a Gauss linear stochastic process with constant coefficients by the static nonlinear transform.then use iterative amplitude matching Fourier method to generate surrogate data.2.Improved time delay vector variance method in the phase space reconstructed, on the basis of different sequence to determine the optimum time delay and the optimum embedding dimension.The geometric feature of the phase space reconstruction and the sequence to be detected are equivalent, they have the same topological structure.3.Calculating the time delay vector variance of the sequence and the surrogate data, draw DVV scatter diagram.Simulation results show that, the stability of improved time delay vector variance is better, the generation of alternative data in the frequency spectrum are more similar to the original sequences; Improved delay vector variance method improves the calculating accuracy of the series, and the practicability is better than the original algorithm.Seismic exploration is mainly divided into three stages:Data acquisition, data processing and interpretation of seismic data.Among them, the data processing is a very important link, through processing the real seismic signal, improving the signal-to-noise ratio, can extract the useful signal effectively and analysis signal characteristics.Improving the signal to noise ratio is closely related to the research of noise properties.And the nonlinearity is an important property of the reflected signal characteristics.Seismic random noise is a noise caused by a lot of ups and downs random harassment accumulated in the time, Seismic random noise’s value cannot be predicted at a given moment. Analysis Properties of noise can effectively remove the random noise, and can effectively preserve the effective signals, so studying on the nonlinearity of random seismic noise has very important practical significance.We can obtain good denoising effect through the research of noise nonlinearity and the parameters adjustment of denoising algorithm. In this paper.using the improved delay vector variance method to process the beginning of the first noise of the desert area, mountainous area, and woodland area, research on the nonlinearity of seismic noise in different areas, Statistic nonlinearity of random seismic noises in different regions.Multichannel actual seismic statistical records show that the desert area seismic noise has good linearity, mountain area and forest area seismic noise has poor linearity. After preliminary analysis,we can get the conclusion that differences of seismic noise nonlinearity in different areas related to the environment, such as climate, topography, geological conditions,they all have different effects on the random exploration noise.