Research On Human Motion Intention Perception Method Of Underwater Booster Robot

Underwater booster robot is a kind of portable foot robot which is used to enhance human strength,underwater booster robot can enhance the movement power of underwater workers,break through the limit of human body,and complete the underwater work that ordinary people can't do.At present,the underwater booster robot can only drive the movement of the lower limbs of the human body in a fixed curve mode,and can't actively perceive the motion intention of the human body,which greatly reduces the practicability and assistance effect of the underwater booster robot.Therefore,this paper analyzes the motion law of the human body in the water,designs and realizes the motion information acquisition and processing platform,studies the recognition algorithm of human underwater motion and lower limb wiggler phase,develop and verify the prediction algorithm of human underwater motion intention,which solves the theoretical and technical problems of perception of human motion intention of underwater exoskeleton robot.This paper mainly studies the perception method of human motion intention of underwater booster robot.First of all,analyzing the law of human underwater motion,summarize the characteristics of human lower limb motion,and get the theoretical curve of the angle of lower limb joint;based on the law of human underwater motion,lay out the sensor module,design the hardware circuit,and build the platform for human underwater motion information collection and processing;based on the platform,complete the underwater motion information collection experiment,and get the original motion data.Then,aiming at the problem of perception of human underwater motion intention,a method based on threshold discrimination is proposed to recognize two kinds of underwater motion states of human body,and an algorithm based on support vector machine(SVM)is proposed to recognize the four lower limb wiggler phase: raising the upper leg up and raising the lower leg up,pressing the upper leg down and raising the lower leg up,raising the upper leg up and pressing the lower leg down,pressing the upper leg down and pressing the lower leg down,and the experiment is designed to verify the algorithm.Experimental results show that the proposed algorithm can accurately distinguish the phase of the lower limb wiggler and the movement intentionof the wearer.Finally,the paper mainly focuses on the prediction method of underwater motion intention,the variable gain Kalman algorithm is used to calculate the angle of human lower limb joint,establishing the prediction model of lower limb joint angle based on support vector regression(SVR)and autoregressive integrated moving average model(ARIMA)to forecast the calculated value of joint angle,and analyzing the prediction effect;proposing the prediction model of lower limb joint angle based on SVR+ARIMA fusion model,and designing the experiment to verify the proposed algorithm.Experiments show that the proposed model can accurately and rapidly predict the next joint angle based on the current joint angle of the lower limb,and achieve the prediction of human motion intention.