Collision Force Estimation And Strategy Research Based On Human-Machine Collaboration Scenario

In recent years,with the rapid development of new technologies and processes in China’s automotive manufacturing industry,many lightweight and flexible components have been used in automobiles.To cope with this situation,human-robot collaboration is required to complete complex assembly tasks.Ensuring the safety of operators and giving the robotic arm the ability to sense collisions and respond to intentional or unintentional physical contact has become a key research issue in the field of human-robot safety interaction.Therefore,this paper focuses on the research of self-sensing collision detection and response technology using an industrial six-degree-of-freedom robotic arm.The accurate dynamic model of the robotic arm is the theoretical foundation for collision detection and compliant control algorithms,and it directly affects the control effect of the robotic arm.Therefore,in this paper,the Newton-Euler method is used to model the robotic arm dynamics,including a linear friction model,and the complete dynamic model is linearized.To simplify the calculation of the excitation trajectory and obtain the optimal value,a fifth-order finite Fourier series is used as the excitation trajectory,and the minimum observation matrix condition number is proposed as the optimization objective,with the robotic arm joint limit and position parameters as constraints.The particle gray wolf optimization algorithm is used to obtain the optimal solution for the excitation trajectory parameters.The mechanical arm dynamic inertia parameters are obtained by using weighted least squares to solve the linear regression equation.Next,a generalized momentum perturbation observer is constructed to achieve the basic function of collision detection.In response to the problem that the GMO algorithm depends on the accuracy of dynamic modeling and is prone to false alarms or omissions in the case of inaccurate modeling,this paper proposes a new collision detection algorithm based on a nonlinear discrete-time self-disturbance rejection extended state observer improved by the fal function.This algorithm eliminates the influence of internal disturbances caused by modeling errors on the observation results,enhancing the sensitivity and robustness of the collision detection algorithm.Simulation experiments are conducted to verify the effectiveness of the algorithm,even in the case of model inaccuracy when the load is less than the rated load.Afterwards,the external force compliant tracking control under position mode is realized by improving the external force compliant tracking algorithm of the perturbation observer based on the fal function.Finally,in order to verify the algorithm proposed in this paper,an experimental platform was established and validation experiments were conducted.The experimental results show that the estimation errors are within a reasonable range,which demonstrates the feasibility of the algorithm proposed in this paper.