Time-varying Correlation Parameter Estimation And Its Application In EEG Signal Processing

EEG signals contain a large amount of information related to physiological diseases,which can be used to diagnose brain diseases and explore brain physiological activities.With the development of science and technology,EEG analysis has entered the computer era,and the correlation research of EEG signals has become a hot spot in EEG signal analysis.Correlation parameters can be used as an important indicator to dynamically evaluate EEG signals and measure brain health situation.According to the traditional analysis method of EEG signals,the long correlation can be measured by Hurst parameter,which is a constant representing the overall self-correlation of EEG signals.However,due to physical functions and external factors,EEG signals in different periods obviously have different local correlation characteristics.Therefore,this thesis gives an improved exponentially weighted H(?)lder parameter estimator for the analysis of local correlation characteristics of EEG signals,and uses the improved algorithm to analyze the local dynamic changes of EEG time series.Based on the analysis and evaluation of the effectiveness of the four traditional Hurst estimators,the concept of H(?)lder parameter H(t)is introduced,and an improved exponentially weighted H(?)lder parameter estimator is given.The parameters and performance of the improved exponentially weighted H(?)lder parameter estimator are analyzed.Simulation analysis shows that the improved exponentially-weighted H(?)lder parameter estimator is improved in accuracy,stability and robustness compared with the traditional H(?)lder parameter estimator.In order to make a better dynamic analysis of the local correlation of EEG signals,this study applies the improved exponentially weighted H(?)lder parameter estimator to two types of EEG signals,epilepsy EEG signals and sleep signals,to dynamically analyze and evaluate the EEG signals.The experimental results show that the improved index-weighted H(?)lder parameter estimator can effectively analyze the local correlation of different types of EEG signals,and provide a novel and effective analysis method for the diagnosis of brain diseases and the dynamic evaluation of sleep EEG signals.