Data-Driven Tolerance Analysis for Products Assembly

Tolerance characteristics have an important influence on products dynamic properties. To realize dynamic properties oriented tolerance analysis and design of mechanical products, a data-driven based tolerance analysis method was proposed. Firstly, based on the cloud model theory, an assemble cloud model was established, and a new method of assembly data processing and knowledge mining was proposed. Secondly, the assembly cloud model was used to analysis the measured data from practical assembly process and search the alignment laws. Finally, using the above mentioned methods, product key parts tolerance was analyzed and adjusted, and the product quality consistency could be improved dramatically.