Research On Status Recognition And Control Of Soft Reptile Robot

The soft robot is a new kind of robot which is almost made of flexible materials.In this paper,a kind of six-legged soft reptile robot is studied,its feet ard made of silicone and body is made of 3D printing,this robot is pneumatic.This paper mainly studies the sensing,state recognition and control of this kind of soft reptile robot,We set up a relatively perfect sensor system,and we have optimized the control of the software reptile robot and verified it experimentally based on getting the message of the state of soft reptile robot.According to the information required for the state recognition of the software reptile robot,the sensor suitable for flexible structure is studied.And we have to consider whether sensors affect the performance of the robot body becauese of characteristics of robot materials.So we compared two kind of conductive silicone rubber and it was determined that conductive rubber was suitable.And we created the multisensor system which prepares for the subsequent robot state recognition.Because of the complex relation between the state information and sensor data of the soft reptile robot,The attitude information of the soft reptile robot was identified by multi-sensor neural network;Environment of road recognition use digital image processing and pattern recognition technology.pavement image is getting by using the visual sensor and we extracted the road surface feature information and established the neural network classifier which can accurately identify four kinds of road surfaces,like sand,grass,gravel and land surface.The method of integrating GPS and inertial navigation is used to identify the motion information of the soft reptile robot.The control method of software reptile robot is based on sequential decision making.According to the state information of the robot identified above,the control parameters of the robot are adjusted so that the robot can walk quickly and steadily.Control parameters are: robot movement gait,inflation pressure and the swing frequency,the evaluation parameters for the robot are speed and efficiency.The optimization goal is to achieve the maximum speed while meeting the motion efficiency of more than 75%.The control system of the soft reptile robot was established,and the simulated sand table of the moving environment of the robot was built,and the feasibility of the control method is verified.