Research On Robot Collision Detection And Compliance Control Strategies Based On Disturbance Observer

In today's society,labor costs are higher and higher,and robots cannot completely replace manpower for production at this stage,especially in some refined operations,the artificial production mode is dominant.In order to reduce labor intensity and improve production efficiency,the production mode of man-machine collaboration emerged.How to ensure the close and safe coexistence between human and robot is an urgent problem to be solved in this mode,and it is also a research hotspot in the field of robots in recent years.Therefore,this paper takes the self-developed 7-DOF robot as the research object,and studies the key technologies such as collision detection and compliance control.The robot dynamics model is the theoretical basis of collision detection and compliance control algorithms,and its modeling accuracy will directly affect the control effect of robot.Therefore,this paper firstly studied the robot dynamics modeling and parameter identification,established the linkage dynamics by Newton Euler method,and analyzed the factors affecting the model accuracy.To reduce the solving difficulty of excitation trajectory,an improved fifth-order finite Fourier series trajectory was proposed to reduce the trajectory parameter from 11 n to 8n,which was solved by the coral reef optimization algorithm based on empirical evolution,and the dynamic parameters were identified by the least square method.Secondly,the collision detection algorithm was constructed by momentum observer and extended state observer,which was improved based on current feedback.Aiming at the problem that the existing collision detection algorithm is easy to fail under the condition of inaccurate model,drawing on the concept of “interference cancellation” in the communication field,a collision detection algorithm based on self-disturbance cancellation extended state observer was proposed,established the relationship between the output torque of the controller and the actual driving torque of the robot in the event of collision by dynamics,eliminated the influence of the modeling error on the observation results of the external force by using the delay response characteristics of the system,obtained the accurate external force estimation formula which is not affected by the size of the load by theoretical derivation,and studied the effectiveness and applicable conditions of the algorithm through experiments.The results showed that the algorithm is suitable for collision detection when the load is less than the rated load and the model is inaccurate.Then,the method of high-pass filtering of the observation results was used to realize the identification of the collision type,studied the collision response strategy,and proposed an impedance control algorithm based on position control and an external force compliance tracking algorithm based on disturbance observer to solve the problem of inconvenient switching of servo control mode.The impedance control algorithm works after accidental collision of robot,so that the robot joint has a certain degree of flexibility,so as to avoid the damage of collision force.The external force compliance tracking algorithm works when the intentional collision is caused by man-made,the observation result of the disturbance observer is mapped into a trajectory through the damping mapping operator,so that the robot can follow the operator's intention to move.Finally,a verification experiment was designed for the algorithm proposed in this paper,the experimental data were obtained through the self-developed 7-DOF robot experimental platform,and the experimental results were analyzed and discussed.The experimental results showed that the detection sensitivity of the proposed collision detection algorithm is more than 2.5 times higher than that of the traditional algorithm.Therefore,the proposed collision type recognition algorithm and the compliant control algorithm based on position control are quite effective.