Stress Assessment And Endogenous Regulation Based On Multiple Physiological Signals

Although it is beneficial for people to have certain stress under stress events,long-term accumulation will inevitably lead to many health problems,which can threaten their lives.Therefore,it is particularly important to accurately identify the stress state and physiological feedback training.In this paper,based on scientific data analysis method to achieve the extraction and optimization of key physiological signal features,and combined with intelligent learning algorithm to build human stress assessment model.Furthermore,the closed-loop training framework of human endogenous regulation and physiological signal analysis is explored,and the training mode of human endogenous physiological regulation with feedback regulation is introduced.By analyzing the influence of regulation rhythm on human stress ability,the effective regulation strategies under stress disorder are evaluated to achieve scientific and efficient training.In the aspect of stress state evaluation and recognition,an experimental paradigm was designed to simulate stress events.The cardiac and brain electrophysiological signals of30 subjects under calm and stress were collected,and the time-frequency domain features of EEG specific frequency band and heart rate variability of ECG signals were extracted.The feature optimization algorithm was used to select 18 dimensional features to form the feature matrix,Then the multi-source feature set is constructed.Using support vector machine(SVM),decision tree,gradient lifting tree and random forest algorithm to evaluate the stress state,the results show that the average prediction accuracy of SVM is the highest,reaching 0.90,and the algorithm accuracy is better than single ECG and EEG,and better than k-nearest neighbor algorithm.In order to evaluate the generalization ability of the algorithm,test data sets are used to verify the effectiveness of the multi-source feature algorithm.In the aspect of feedback regulation under stress,this paper proposes an endogenous regulation strategy of stress disorder under stress events,adopts a closed-loop training framework combining human respiratory regulation and physiological signal analysis technology,explores the influence of different respiratory regulation modes(rhythm,depth,etc.)on human stress ability,and puts forward practical respiratory regulation methods.And further simulate the stress events,using the "inter subject design" method,the samples were divided into different modes of respiratory regulation group and control group,different respiratory regulation groups were divided into quiet period,stress period and corresponding respiratory regulation period.The control group was in the period of calm,stress and self recovery.Based on the statistical method,the data of each group were analyzed scientifically.The experimental results showed that during the recovery period of each group,the physiological indexes of the subjects in the respiratory control group recovered to the calm period within a certain period of time,while the control group did not recover to the calm period within the same period of time,which proved the effectiveness of the respiratory control.Furthermore,according to the time domain analysis,the stress recovery time of the respiratory control group was shortened by more than 33% compared with the natural recovery time.The results show that the endogenous regulation method proposed in this paper can be used as a feedback training method under stress disorder.