Research And Implementation Of External Forces Detection Of Industrial Robots For Human-robot Collaboration

With the official promulgation of the Planning for "Made in China 2025","Intelligent Manufacturing" will become the future development trend of the manufacturing industry.In the traditional manufacturing mode,industrial robots(IRs)have been widely used in some scenes with high repeatability and requiring strict work space isolation of human and IRs.However,to utilize the dual feature advantages of robot's good repeatability and human's high flexibility in certain production scenes of the manufacturing industry,the Human-Robot Collaborative Manufacturing(HRCM)model emerged,which has attracted great interest in the world's robotics field.In terms of the research on HRCM,the issue of Human-Robot security is of most importance.Detecting the accidental collision forces in HRC and controlling it in real time is of great significance to Human-Robot security.In this paper,aiming at the poor real-time performance of the existing methods,a real-time external forces detection method for industrial robots based on dynamics model is proposed.Meanwhile,Aiming at the shortcomings of high false alarm rate for external forces detection methods using static thresholds,an adaptive threshold based algorithm is proposed for dynamic external forces detection.The main research contents are as follows.(1)Research and implementation of external forces detection method for IRs based on dynamics model.Based on a real-time communication channel between the PC and KRC,a real-time dynamics robot model based method of external forces detection is proposed.A comprehensive experiment is performed on parameter identification process,including: establishment of a linear dynamics model,design of optimized joint excitation trajectories,acquisition of experimental data and so on.Finally,the established dynamics model is applied to the external forces detection method.An experimental platform is built to verify that the parameters identification result has higher accuracy under certain conditions and the proposed method can initially realize the real-time external forces detection for IRs.(2)Research and implementation of a dynamic method of external force detection for IRs based adaptive thresholds.After a detailed derivation of the residual dynamics model,on the basis of the dynamics model based method of residual torque calculation,an adaptive external forces detection algorithm for IRs is proposed.In order to overcome the residual noise caused by static uncertainties such as load change,a high-pass filtering operator for residual signals is first designed.To overcome the residual's dependence on robot's real-time kinematic parameters,a dynamic thresholds synthesis algorithm are designed according the input information on robot kinematic parameters and the past output results of dynamic thresholds.Besides,a detailed adaptive threshold-based residual comparison strategy is formulated,and the unknown parameters in the algorithm are estimated using a recursive least squares method to update the dynamic thresholds in real time.Finally,the experimental platform is built and certain related experiments are designed to verify the performance of the algorithm.After comparing with other external force detection algorithms,it's verified that the performance of the proposed algorithm has got some improvement on certain performance indicators,such as false warning rate.(3)Design and implementation of external forces detection and safety control system for IRs.In order to verify the practicability of the proposed external force detection method in this paper,a set of external force detection and safety control system is designed and implemented.While realizing the function of dynamic external forces detection,the robot admittance control module is designed and developed to realize the safety response control for IRs.The detailed implementation process is given for the external torque estimation module,the external force detection module and the robot admittance control module of KUKA industrial robot.Finally,a comprehensive experiment is designed and conducted on the external forces detection and admittance control for the IRs.The experimental results confirm the effectiveness of the adaptive threshold based method of external forces detection and the successful execution of admittance control for KUKA industrial robot according to the detected external forces.