Automatic Assembly of Eccentric Shaft Component Based on the Machine Learning Algorithm

An assembly strategy based on the machine learning algorithm is proposed to cope with the difficulties in the practical assembly of eccentric components in high-precision RV speed reducer. The simplified models of eccentric shaft hole are built to analyze the possible contact states in the process of assembly. Subsequently, the basic principles of the machine learning algorithm are introduced, and a shaft hole assembly method with supervised learning is put forward. At last, the platform for assembly robot is constructed and tested with force characteristic information, and the test results indicate the good effect of this assembly strategy.