Robot-assisted endoscopic minimally invasive surgery has attracted wide attention due to its characteristics such as small trauma,low infection rate and quick postoperative recovery.The related technologies of endoscopic surgical robot are also gradually becoming the hot topic in the field of robotic research.For those reasons,some key technologies of endoscopic surgical robot are studied,mainly involving the mechanism design and optimization of the haptic master manipulator,the master-slave motion control of endoscopic minimally invasive surgical robot,the interaction force detection model of endoscopic minimally invasive surgical robot,the physical human-robot interaction of the master manipulator as well as the control strategy of the master-slave force-feedback.A fully active serial master manipulator is designed for the endoscopic minimally invasive surgical robotic system.The parallelogram mechanism and the position compensation mechanism are used in the master manipulator to realize the independent adjustment of master manipulator’s position and attitude.In this way,the comfort and intuition of the doctor’s operation are enhanced and the complexity of kinematic analysis and calculation is reduced.Based on the special application scenarios of minimally invasive surgery,a mechanism optimization index that considers dexterity,positioning accuracy and structure length utilization is proposed to better optimize the mechanical settings and improve the comprehensive performance of the proposed manipulator.In order to obtain the optimal solution of the optimization index,an improved particle swarm optimization algorithm with penalty term is proposed,which can improve the global search ability of the optimization algorithm and avoid precocious convergence.According to the functional requirements of minimally invasive surgery,a reasonable master-slave control strategy and a master-slave intuitive motion mapping model are established.Aiming at the motion error caused by the nonlinearity of the cable-pulley mechanism of minimally invasive surgical instrument,the master-slave motion control strategy is improved.The proposed schemes based on support vector machine and feedforward neural network can use the time domain and frequency domain information of the driving motor current and the motion information of surgical instrument to identify the movement stage of the end-effector online and in real-time.The positioning error caused by backlash hysteresis and coupled motion of the cable-pulley mechanism is analyzed based on the movement stage of the end-effector and eliminated by the feedforward compensation and then the master-slave control accuracy can be improved.In order to accurately obtain the contact force between the instrument and the tissue during minimally invasive surgery,an interaction forces detection system based on 6 axis Force/Torque sensor is designed.A dynamic model based on the virtual axis is built to analyze the relationship between the measured force information,the interaction forces and the additional forces produced by the motion of the instrument.In order to accurately obtain the dynamic parameters,firstly,a data processing method based on gaussian process regression and zero-phase lowpass filtering is designed.Then the excitation trajectory model is established and the trajectory parameters are optimized.The dynamic parameters of the surgical instrument are identified by the least square method.Finally,a detection scheme which can obtain the contact force between the surgical instrument and the tissue in real time by using the position,velocity of the surgical arm’s active joints and the measured 6 axis force information is proposed.In order to avoid the poor compliance of the master manipulator affecting the operation experience and the feedback-force perception of the doctor,a physical human-robot interaction control strategy of the master manipulator is proposed.A real-time observer based on generalized momentum is established and the time delay neural network is used to compensate the calculation error of the observer model.The compensator based on time delay neural network can further improve the compliance of operation.The proposed shceme has implemented the active compensation of the additional torque of the master manipulator during the operation and can adjust the degree of operation compliance according to the operator’s requirements.Based on the proposed human-robot interaction control strategy of the master manipulator,a master-slave force-feedback control strategy is developed.The force-feedback control is conducted by using the designed fully active master manipulator according to the detected interaction forces.It is guaranteed that the operator can feel the interaction force information of the slave side at the master side and the direction of the feedback-force is consistent with the direction of the interaction force in the 3D video image.On the basis of the above studies,the experimental platform of robot-assisted endoscopic minimally invasive surgery is established.The master-slave intuitive operation verification experiment,the nonlinear compensation experiment of cable-pulley mechanism,the tool-tissue interaction forces detection experiment,the physical human-robot interaction experiment of the master manipulator and the tissue palpation experiment in the master-slave mode are carried out.The experimental results show that the study of this paper can ensure the intuitiveness of master-slave operation,effectively reduce the nonlinear error of the cable-pulley transmission mechanism,improve the master-slave control accuracy,realize the detection of the interaction forces for endoscopic minimally invasive surgical robot and feedback the detected interaction forces to the operator by using the designed master manipulator. |