Design And Application Of Mental Fatigue Detection System Using Non-Contact ECG Measurement

As the pace of life accelerates,people's work pressure increases and sleep time decreases.Mental fatigue has become an important issue.When a person is in a state of mental fatigue,they are drowsy,unresponsive,and inefficient,which may further cause an accident.Accidents caused by fatigue in some special populations can be extremely significant,such as drivers,doctors,and dangerous instrument operators.Research on mental fatigue monitoring has a long history,but most of them need to be equipped with complex testing equipment.Therefore,it is very necessary to provide a set of undisturbed and portable high-precision mental fatigue monitoring system.This paper designs a non-disturbing mental fatigue monitoring system.In this paper,the capacitive coupling electrodes are used to collect the electrocardiogram(ECG),and the piezoelectric sensor and its conditioning circuit are used to collect the ballistocardiogram(BCG).Through the analysis and modeling of the data in the mental fatigue experiment,a model for evaluating mental fatigue was established.Software monitors and alerts based on models.In the design of this paper,the acquisition circuit of coupled ECG and BCG was designed first.Then digital signal processing algorithms for physiological signals are designed on tablets and desktops,including de-baseline,bandpass filtering,adaptive filtering,fast Fourier transform,etc.The data collected by the system is verified by standard medical signal sources,patient simulators and heart rate bands.This paper designed the mental fatigue experiment.The data acquisition system collected data of five subjects in the mental fatigue experiment.Through the time-frequency domain analysis of these data,a mental fatigue determination method based on heart rate variability(HRV)and heart rate is obtained.In addition,in this paper,the machine learning model is established for these data.In the support vector machine,the Gaussian kernel function is used to classify the data features into three categories,which represent the normal,fatigue,and drowsy mental states of three different degrees.The accuracy rate is reached.93.3%.The innovation of this paper is to collect the coupled ECG and ballistocardiogram at the same time,monitor the mental fatigue through non-contact heart rate variability parameters without disturbance,and support multi-user management of remote terminals.Compared with other similar studies,this system has certain advantages in comfort,reliability,accuracy and practicality.