State Perception And Safety Control Of Grinding Process In Spine Surgery Robot

As a branch of orthopedics,spinal surgery is a high-tech,high-risk subjects.During the operation,doctors need to complete complex movements in a small space,and it will increase the risk of surgery or even makes the operation failed because of fatigue or lack of vision and other reasons.Robot has high positioning accuracy,high stability and other characteristics,which can improve the traditional operation mode and assist the doctor to complete the operation.In terms of vertebral lamina grinding,there are few related researches on the use of robot assisted to complete the vertebral plate grinding operation,and many key technologies are not mature enough and need to be further explored.In this paper,we take laminar milling as the research background to study the related field of the spinal surgery robot system.Including the research of registration algorithm and automatic registration,the state sensing based on current and the variable feed grinding,the safety control based on virtual fixture and the lamina grinding experiment.In this thesis,we will study the registration algorithm and discuss the existing algorithms.Aiming at the shortcomings of the existing algorithm,such as low precision and high computational cost,a new registration algorithm named TF(Tetrahedral Framework registration algorithm)is proposed,which uses three-point registration,multi-point optimization.The process is easy to understand,and it has small amount of calculation.In order to achieve efficient registration,this paper will introduce an automatic registration device based on marker point to realize automatic registration of navigation system.In order to simulate the feel of doctor's hand,this paper will study the perception of the state of the robot,and based on the current signal,through the wavelet transform,changing rate threshold to mutation,normalization and so on,to complete the state of bone identification,so as to achieve the perception of the state of the robot.In this thesis,we will discuss the variable feed grinding based on state perception.Increasing the feed rate when there is no contact with the bone,and Decreasing the feed rate when there is contact with the bone,then that variable feed grinding method can achieve safe and efficient grinding.In order to realize the precise guidance on the grinding path of the robot,this paper will propose guided virtual fixture with admittance factor and damping type virtual fixture based on position compensation,which can achieve the curve tracking and blocking the robot to enter a specific area respectively.Finally,this paper will combine the technology of robot navigation system,state sensing and motion constraints to achieve the experiment of lamina milling with safety and high efficiency.