The Measurement Of Driving Fatigue State Based On Soft Set Theory

Surface electromyography(sEMG)are the biological electrical signals,which are generated by nerves and muscles when people perform autonomic movement.The signals can reflect the physiological activity and functional status of people,and have been widely applied in various fields,including clinical medicine,man-machine interaction,ergonomics and fatigue diagnosis.This thesis focused on the feature extraction and the fatigue analysis of the electromyographical signals in driving state.The fatigue quantitative model based on the soft set theory is established to assess and quantify the state of fatigue effectively.The contributions of this thesis are as follows:Firstly,the single physiological signals have the low accuracy and are easily disturbed in the process of fatigue detection.In order to overcome these problems,multiple muscles of subjects are collected.Two methods,wavelet de-noising method and the EMD reconstructing de-noising method are used to deal with multiple signals.Secondly,the clearer separation degree and characteristic parameters are achieved by analyzing the time domain,the frequency domain and the complexity of entropy,which can better reflect the human state.The approximate entropy,sample entropy and fuzzy approximate entropy are discussed in processing the sEMG respectively.Then the results show that the fuzzy approximate entropy method has some advantages,such as better relevance,consistency with the degrees of muscle fatigue,little dependence on parameters and stronger robustness to noises,which can better reflect the changes of fatigue state of the sEMG.At last,the fatigue measurement model based on the soft set theory is established.The decision-making idea and the abnormal data processing technique to physiological signal processing field are introduced.The characteristic values of the sEMG from multiple muscles are processed by two layers of data fusion.The processing abnormal data algorithm is improved.According to different fatigue sensitivities of the sEMG corresponding with different muscles,the weights of parameters are given and the final decision result is more stable.In order to verify the effectiveness of the proposed model,the RBF network method is discussed.The results show that the state information given by the model based on the soft set theory is more detailed and more readable.