Fatigue Identification For Random Driver Based On HMM

Driver fatigue is one of the main causes of serious traffic accidents.If the fatigue state can be accurately identified in real time and the driver can be timely warned,the traffic accidents can be effectively reduced.At present,relevant scholars at home and abroad have conducted a lot of research on the identification of driver fatigue,and have achieved a series of research results.However,there is still a lack of in-depth systematic research on the intrinsic relationship between individual characteristics of drivers and the performance of recognition systems.Based on this,this paper aims to improve the adaptability of the fatigue recognition system.Based on the analysis of the difference of individual characteristics of drivers,this paper proposes a HMM-based fatigue identification for random driver,and the model is validated by relevant experiments.The main research contents include:(1)The mechanism of fatigue identification for random drivers.Based on the analysis of the difference of driver’s individual characteristics and the influence of driver’s individual characteristics on driver fatigue identification system,a method of eliminating the difference of driver’s individual characteristics based on model parameter transformation is proposed.Then,the specific fatigue identification models of drivers with similar individual characteristics to the tested driver are selected,and the fatigue identification model of the tested driver is obtained by linear weighted summation and adaptive operation of the parameters of each specific model,which realizes the identification of the fatigue state of the random driver.(2)The construction of the basic model of driver fatigue identification based on HMM.Based on the analysis of the research status of driver fatigue identification,this paper selects EEG signals which can objectively reflect the changes of driver’s state as feature data,and uses wavelet transform to complete the extraction of feature parameters,and then constructs a basic model of driver fatigue identification model based on the HMM with dynamic characteristics.Finally,the basic model is validated based on experimental data.(3)The construction of the fatigue identification model for random driver based on HMM.Firstly,in the model construction stage,based on the analysis results of the individual characteristics of the driver,the FCM algorithm is used to classify the driver,and then the specific fatigue identification model is constructed based on the individual characteristics of each classification.Secondly,in the state identification stage,the corresponding specific identification models are determined based on the individual characteristics of the testeddriver,and then the model parameters of the tested driver are determined by the proposed model parameter adaptive algorithm.Finally,based on the experimental data,the fatigue identification model for random driver constructed above is tested and validated.Based on the correlation analysis between the individual characteristics of the driver and the fatigue recognition system,this paper proposes a fatigue identification method for random driver based on the model parameter adaptive method.The test results show that the proposed method can effectively improve the adaptability of drivers with different driving characteristics under the premise of ensuring the recognition accuracy.And the relevant research results have practical theoretical implications for the further performance improvement of the vehicle system.