Wrist Multi-parameter Injury Monitoring Device

According to statistics,90% of the soldiers killed in the battlefield are killed because the injury is not treated in time.So it is especially important to ensure the safety of the injured soldiers in the wild.Traditional medical equipment is bulky and has high power consumption,which cannot meet the needs of field use,and the security of collected sign data cannot be guaranteed.In view of the above situation,this paper designs a multi-functional and miniaturized wearable wrist multi-parameter injury monitoring device,the main work is as follows:(1)Design a wrist-type wearable device for detecting the three parameters of oxygen saturation of blood,heart rate and blood pressure,and transmit data through near-field communication to meet the requirements of miniaturization and low power consumption of the device.The device consists of a oxygen saturation of blood and heart rate module based on the MAX30102,a blood pressure module based on the MPS3117 pressure sensor,a core processing module based on the STM32F103RCT6,an NFC module based on the RF430CL330 H,and an OLED display module.(2)The pulse wave signal collected by the device is smoothed,the baseline drift are removed and the peak of the signal is detected,which lays a foundation for accurate calculation of the three parameters.In this paper,the Fluke Prosim8 multi-parameter vital sign simulator is used to calibrate and fit the parameters of the calculation formula of blood oxygen and blood pressure.The accuracy of blood oxygen calculation is about ±1%.The average error of systolic blood pressure measurement is 2.56 mmHg,and the standard deviation is 3.68 mmHg.The average diastolic pressure error is-6.27 mmHg,and the standard deviation is 3.10 mmHg,which can meet the standard of accurate measurement.(3)Low power design optimization was performed from the system level to the device.NFC module works with the card emulation mode which power consumption is the lowest.The core processing module combines the stop mode provided by the STM32 CortexM3 core to adjust the clock and voltage in a timely and dynamic manner in the device measurement workflow.The oximetry and heart rate module adjusts the sampling rate by feedback of the measurement results of the sign parameters,and sets the reflected light threshold to detect when the device is not worn to reduce the current and reduce the power consumption.After testing,the core processing module reduces the current from 93.6mA to 53.0~53.2mA during the measurement interval.The current of the oximetry and heart rate module is maintained at 2.1~2.2mA when not worn.The operating current of the 400 Hz sampling rate is 8.22~8.24 mA.The 100 Hz sampling rate reduces the operating current to 3.8~4.5mA.(4)In order to improve the security of the collected data in this paper,the frame format of the sign data transmission with NDEF as the carrier is designed,and the accuracy and integrity of NFC data communication are verified.Furthermore,the feasibility of the application of RF fingerprinting technology in this device system is studied and analyzed.In this paper,the RF fingerprint features of the device are extracted by wavelet transform and bispectrum estimation.Then,using the support vector machine through the grid search and cross-optimization method to train and test the model of four labels of 400 sets of feature data in a ratio of 4:1.In the process of parameter optimization,the accuracy of model identification is in the range of 85%~99.375%.The model is trained by the parameter value of the optimization result.The recognition accuracy of the test set can reach 92.5%,which can verify that RF fingerprint identification technology has certain feasibility in this system.