Research On The Control Strategy For The Flexible-joint Robot

With the rapid development of robot technology,the application field of robot has also made a large expansion.The traditional industrial robots are developing gradually to the new fields and service industries which are used to assist human activity at present,so the work space of robot is required to transfer from the structured and protected fixed work space to unstructured and interactive collaborative work space.The robot which has security physical interaction ability becomes the inevitable trend for the robot development.And the flexible-joint robot is one of the effective ways to realize the security interaction.Therefore,currently the research on the control strategy for the flexible-joint robot has become one of the important fields of the robot research.This paper will take the flexible-joint robot as the research object,and study the control strategy of the flexible-joint robot as the research purpose.The security physical interaction will be achieved between the robot and environment or human beings depending on the control strategy.This paper is supported by the National Natural Sciences Foundation of China titled“Dynamic and compliant control of dual–arm of humanoid robotic”,the project of Beijing Technology Committee titled“Research on press manipulation of TCM massage robot aimed at pain of waist”and the“12^th Five Year”National Science and technology support project,named“Study on humanoid Chinese medicine rehabilitation robot for stroke hemiplegia”.The relevant work about control strategy of the flexible robot is developed,it is very significant to achieve robot security interaction,and it can be used by the other service robots.The main work of this paper includes:1.The dynamic model of the 4-DOF flexible-joint robot is derived using the Lagrange method,and study the coupling which is not only though the inertial matrix,but also via the torque of elastic element between the motor flexible subsystem and the rigid link subsystem.The reduced dynamic model of the robot is established.The effectiveness and rationality of the reduced dynamic model of the flexible joint robot is described,and the simulation study is carried out.2.The control strategy for the flexible joint robot is presented based on hierarchical control structure.The robot system can be divided into two subsystems which have different time scale depend on the singular perturbation principle,and the stability is proved by the introduced approach.The system stability is verified by computer simulation.And analyze the system robustness of the flexible joint robot without precise gravity compensation and accurate elastic element stiffness.3.For the proposed hierarchical control structure,the possible peaking phenomenon of the flexible joint robot is analyzed depending on the high gain of the torque loop,and the effect of peaking phenomenon on robot stability is studied.4.Because the dynamic parameters of flexible-joint robot can't be measured accurately,the gravity torque can't be calculated directly by the gravity torque expression.The neural network approximation is presented to achieve gravity torque calculation,and the effectiveness of approximation is verified by experiments.Finally,the influence of the precise gravity compensation on the system is analyzed.5.The impedance control method of the flexible-joint robot is presented based on the hierarchical control structure.Introduce the relationship between the system stiffness and control parameters based on the different position sensors configuration.Finally,the system stiffness of the flexible-joint robot based on the position feedback of the links is verified which can be more rigid than the elastic element of the joints.6.Two experimental platforms are established including the flexible joint and flexible-joint robot with four degree of freedom,and all experiment studies are carried on these platforms.The concrete research contest includes the following several aspects:the experimental research on robotic feature;single joint position and impedance control experiments;robot position and impedance control experiments.The experimental results show that the proposed control strategies can effectively control the flexible joint and the flexible-joint robot.