Design Of Plantar Pressure And Ankle Joint Bending Sensor And Research On Gait Analysis System

Walking is a process of coordinated movement of the human trunk and multiple joints.Abnormalities in any link may lead to abnormal gait.Through gait analysis,it can be judged whether the gait is abnormal and the reason for the abnormality.At present,the widely used gait analysis system equipment is complicated and only exists in hospitals and laboratories.So,it is difficult to promote it on a large scale.To this end,this paper designs a flexible plantar pressure sensor and ankle joint bending sensor based on liquid metal,and uses these two sensors to design a wearable gait analysis system that can provide real-time monitoring of the wearer’s plantar pressure and ankle bending angle were measured and their gait was analyzed.The main contents of this paper are as follows:(1)Choosing silicon rubber as the base material of the sensor and EGa In liquid metal as the sensor’s sensitive element,developed the plantar pressure sensor and the ankle joint bending sensor.The sensor’s structural design and manufacturing process research have been completed.The test method is used to determine the manufacturing parameters of the liquid metal sensor,including the hardness of the silicone rubber,the height of the sensor and the size of the micro-channel.It is proved that the plantar pressure sensor and the ankle joint bending sensor have good repeatability.(2)The gait data acquisition system is developed with the plantar pressure sensor and the ankle joint bending sensor as the measurement unit.The system hardware uses the Arduino Mage2560 micro-control board as the control module.The signal amplification module,the wireless transmission module and the power supply module are also designed.Using Savitzky-Golay sliding filter filtered the signal.Completed the periodic segmentation of the gait data.And analyzed the change characteristics of the pressure in each area of the plantar and the bending angle of the ankle in a cycle.(3)A gait period division algorithm based on support vector machines is proposed.The gait data is feature extraction and normalization.The Gaussian kernel function is selected using the cross-validation method.The Gaussian kernel function is separated from the linear kernel function.It is mixed with the Sigmoid kernel function to improve the performance of the gait period division model.The particle swarm algorithm is used to optimize the parameters of the model.The collected gait data is used to test the gait period division model,and the test results indicate the gait phase of the model The recognition accuracy rate can reach 96.12%.The division of the gait cycle is completed in the software,and the display of the division result is completed through the interactive interface.