Foot Feature Extraction And Fatigue Detection Based On Plantar Pressure

With the acceleration of people's life rhythm,more and more people lack sufficient physical exercise due to long-term overtime work and irregular lifestyle,which makes the body in a sub-health state for a long time,and strengthening physical exercise is an effective way to promote health.The use of fatigue detection methods to monitor the state of muscle function during exercise can detect muscle damage in the human body in advance and improve the effect of exercise effectively.This thesis designs and produces a set of exercise fatigue recognition system based on eight-way plantar pressure signals,using ZNX-01 pressure sensor to collect plantar pressure signals and MPU9250 chip to collect acceleration signals.Through the analog-to-digital conversion module inside the STM32F103C8T6 control chip,the continuous voltage signal and acceleration signal output are sampled into discrete digital signals,and the processed data is sent to the host computer system via Bluetooth.The upper computer can realize real-time drawing and storage of waveform data,and load the machine learning model that has been trained to determine the level of sports fatigue on the collected plantar pressure data,and output the results of sports fatigue recognition.The system is simple and easy to implement,has strong operability,and has a good application prospect.This thesis demonstrates the advantages and disadvantages of the traditional method of quantifying the level of exercise fatigue through the Borg scale,and proposes a method to quantify exercise fatigue based on the maximum oxygen uptake.The traditional Borg scale method is simple and easy to implement,and has strong operability,but it is highly subjective.Based on this situation,this thesis proposes to mark the collected plantar pressure data by the percentage value of the maximum oxygen uptake.Five exercise fatigue levels of 60%,70%,80%,and 90% of normal state and maximal oxygen uptake are determined.Theoretical research and experimental results show that the method of quantifying exercise fatigue based on maximal oxygen uptake is feasible.The experimental design is divided into static and dynamic plantar pressure measurement experiments.The results of static experiments showed that the pressure on the eight areas of the sole of the foot accounted for the total pressure,the largest is the heel,followed by the first metatarsal,the second to third metatarsals,and the first phalanx has the smallest proportion.The dynamic plantar pressure measurement experiment is a multi-level exercise intensity experiment,which analyzes the collected data from the perspectives of statistics and machine learning.The experimental results show that the fit of statistical analysis is general,while the exercise fatigue recognition model based on the random forest algorithm,the gradient boosting tree algorithm and the support vector machine algorithm has a high prediction accuracy rate,and the recognition accuracy rate of the random forest model reaches 82.5 %.This project also calculates human energy consumption based on the acceleration signal at the ankle.The experimental results show that the change trend of the average energy consumption of the subjects at all levels of exercise intensity is consistent with the change trend of exercise intensity.