Human Physiological Information Detection And Self-Energy System Research

In modern medicine,the high-quality equipment for physiological information detection is scarce,expensive,and mostly imported from abroad.The detection equipment used commonly is low in integration and mostly relies on battery power.The cost of battery replacement is high and it is easy to cause environmental pollution.Based on this situation,this article has carried out research on human physiological information detection and self-powered system,which is powered by piezoelectric self-powered devices.The low-power programming is applied to reduce the energy consumption of physiological information detection system,and the neural network algorithm is used to predict and classify the acquired physiological information.The main research contents of this paper are as follows:Firstly,the overall design scheme of physiological information detection and self-supply system is proposed.This scheme divides the shoe body from top to bottom into a BCG signal detection layer,a gait detection layer,and a self-powered device layer,which are used to detect human resting BCG signals,dynamic pressure signals,and collect and transform human kinetic energy.The feasibility of the research program was demonstrated in combination with related theories such as human motion mechanics,piezoelectric effect and physiological information generation mechanism.Secondly,around the energy balance,the self-powered device and the physiological information detection system were designed separately.In the design of the self-powered device,the piezoelectric technology is combined with the spring amplification technology,and the 3D structure is designed through Rhinoceros and the physical production is completed.At the same time,an improved piezoelectric energy trapping circuit is proposed,and the performance comparison with the LTC3588-1 circuit confirms the effectiveness of the improved circuit.The actual wear test results show that the self-powered device captures about 0.7308 J of energy per unit hour.In the design of the physiological information detection system,the module selection,mode configuration and current test have been completed.In order to reduce the running power consumption of the detection system,a low-power program running scheme was proposed and the power consumption in a single running cycle was calculated to be 58.1m J.From the perspective of energy balance analysis,it can be seen that the use of intermittent data collection and analysis can ensure that the system completes a data interaction process with the host computer within 423.1 seconds.Finally,the research on physiological information detection algorithm and prediction classification is carried out.In the research of Ballistocardiogram signal,the heart rate is extracted through differential threshold and peak detection algorithm,and the t(IJ),t(IK),Int(HK)and Int(BCG)feature vectors in the cardiac shock signal are extracted as the network input.Comparing IPSO-BP and BP prediction models,it is found that the performance of the IPSO-BP classification model is better,and the IPSO-BP neural network is used to classify and identify coronary heart disease.The experimental results show that the overall accuracy of the model is 87%,and the accuracy of coronary heart disease prediction is 92%.In the research of gait detection,the sensor placement points are marked according to the human body dynamic and static states,and the period mean,frequency,period standard deviation,stance reference line and sensor strength characteristic parameters from the pressure waveform were extracted as the network input.The ACO-Elman neural network is used to identify the standing,walking,running and falling states.The experimental results shows that there are differences in experimental populations of different genders,and the discriminant indexes of gait detection exceed 85%.