Assistance with the development of innovative assembly procedures
This work consists in assisting with the choice of an assembly procedure based on several criteria: the component's cost as a function of manufacturing accuracy, the cost of the chosen production apparatus, and the nonconformance rate in the simulated assembly. A method was developed for assisting with the choice of the generic procedure and a software platform was proposed. As part of this work, we proposed an architecture for the assembly data, a cost evaluation method for the components and the manufacturing tools, and a method for the estimation of the conformance rate. In order to achieve an optimal choice of the tolerances and assembly methods, we proposed a mechanical model reduction based on a metamodel using neural networks, and we implemented a multi-objective optimization method.