Soldier Mental Status Monitoring System Based On Wireless Body Area Network

Soldier status monitoring system is an indispensable part of the modern warfare.The current state of the soldier monitoring system usually has few parameters,and most of them can only judge the soldier's vital signs,which leads to its limited application area.With the development of wireless body area network technology and sensor technology,comprehensive analysis of various physiological p arameters of individual soldiers to determine the physiological and mental state has become the hot topic domestic and overseas.Aiming at the problem of the parameters limitation,low communication speed and the low practicality of soldier status monitoring system,we design a multi-parameter soldier monitoring system using Wi-Fi wireless communication,and use distributed nodes to monitor a series of physiological parameters such as EEG,EMG,ECG,Breath and body postures in this paper,and on-line processing and analysis to determine the soldier state based on DSP main node is also implemented.Then we analysised the system accuracy,power consumption,reliability and other propertie objectively.Due to the complexity of physiological signals' characteristics,this paper discusses the proccessing methode of physiological parameters and the method of feature extraction seriatim.The noise signals such as baseline drift,power frequency interference and physiological artifact are eliminated,and the percenta ge of EEG power spectrum density,blink frequency,Stability,respiratory power,posture Euler angle,etc.a total of 20 kinds of physiological signal features has been extracted.The existing soldier monitoring system usually only monitor vital signs,and can not achieve mental state monitoring.For this difficult problem,this paper uses the EEG and blinking frequency to achieve the four classifications of mental fatigue by BP neural network.For the emotional state,this paper divides it into positive,neutral and negative,and adopts the method of multivariate physiological parameter fusion analysis to identify the emotions of specific objects through support vector machine.In this paper,mental fatigue and emotional experiments were designed in the laboratory environment.The various physiological data were collected by using the designed soldier monitoring system,and the validity of the method was verified.Experiments show that the accuracy of the four classification of mental fatigue is up to 83% for a number of different subjects,while the accuracy of emotion recognition is 92.5% for the individual solder.