Fine blanking sprocket medical robot working environment and human survival environment, manipulator will regularly contact with human body and the work environment, ensure the safety of the people is the primary condition, which should be followed. Along with the increasingly complex mechanical arm control, how to realize the mechanical arm, switch the control strategy of fault classification, the premise of a successful action.
Based on rigid and flexible mechanical arm, for example, focus on the momentum of residual fault detection algorithm, based on support vector machine (SVM), the motor fault and collision failure of the design of fault classifier to classify two arms of the most common fault in the operation process. First of all, based on the Lagrange principle, using screw theory and mode, try to establish a dynamic model of rigid and flexible manipulator arm, the deduction in detail. Second, make the mechanical arm, the derived residual fault detection algorithm, the dynamic model of Matlab/Simulink to simulate platform construction.
Again, use two to three degrees of freedom rigid robot arm Adams to build a virtual prototype model of the fault simulation, through Matlab and Adams simulation using Matlab/Simulink/SimMechnics established the rigid and flexible weapons of two - degree of freedom model, simulation of the complete failure. The results show that the residual momentum values derived from different system model, for the same fault with the same characteristics. Then, the principle of fault classifier based on support vector machine (SVM) by two shots with three degrees of freedom of momentum just arm 7 fault data as the residual value of the sample data to extract the feature vector using Libsvm train and test classification and classification results. Finally, staubli RX90 robot experiment, the extract of kinematics just arm data to obtain the momentum of the residual value, extract the feature vector, the classifier training and test, the test results. The momentum of the remnants of failure detection algorithm is extended to the field of the flexible manipulator,
Simulation platform is established, the comparative analysis of rigid and flexible mechanical arm residual momentum value characteristics of experimental verification, the fine blanking sprocket residual momentum rigid manipulator based on support vector machine (SVM) the effectiveness of the fault detection and classification algorithm of fault detection and classification of residual based on momentum provides a basis for practical applications.
