System Design And Human-robot Co-operation Control Of Robot For Craniotomy

Craniotomy is a common approach for treatment of glioma,skull base injury and tumor.However,its shortcomings,for instance,time consumption and high risk,makes it highly dependent on clinicians' skill and judgement.In this thesis,a robot-assisted craniotomy system is proposed in order to improve clinical procedure,with respect of accuracy,transparency and safety.The robot system employs human-machine collaborative control,so as to keep the clinician in the dominated place in the control loop,as well as to take advantage of the sensitivity and precision of the robot.Surgical requirements are discussed carefully,based on which a robot system is devised and a collaborative control method is developed.A series of simulations and experiments demonstrate the proposed approaches.The research is based on projects funded by National Natural Science Foundation of China under grant No.61540036 and Beijing science and technology plan under grant No.Z161100001516011.Main research contents and highlights are as follows:First of all,analysis of surgical requirements and proposal of the robot-assisted craniotomy system.According to multiple rounds of discussions with experienced surgeons and surgery visits,a full requirement summary is completed and based on which,the framework of a parallel human-machine collaborative craniotomy robot system is sketched.The layout in the operation room and the robot system's basic workflow are also discussed;Secondly,modeling and kinematic simulation of the robot arm.A six degrees of freedom(DOFs)robot arm is employed in the robot system.The model of the robot arm is built using Creo and transplanted into Adams,where its kinematic simulation is carried out together with Matlab/Simulink.Differential method is used to perform motion control and Monte Carlo method is used for workspace analysis.Finally,threedimension circular arc interpolation algorithm is adopted to realize irregular curve path planning in space;Thirdly,a parallel collaborative control method is proposed and validated via simulation based on impedance and admittance control theories.With 6-axis force sensors to detect the clinician's intension,this method,built on top of position control,can improve safety and flexibility during the interaction between the clinician and the robot;At the last,the Lagrange energy method is employed for gravity compensation of the load on the robot arm,therefore alleviating the clinician's burden during operation.Dynamic compensation is carried out using the Newton-Euler algorithm,then dynamic decomposition and trajectory tracking of the robot arm can be achieved.