An Optimization Study on Assembly Errors

Products cannot be produced as perfect as nominal values due to errors in manufacturing and assembly processes. These complex errors eventually affect the function and performance of the finished products. This paper deals with how to minimize the overall errors when assembling each part which has the manufacturing errors. We used a genetic algorithm and coded the coordinate transformation including other necessary mechanisms to move objects in three-dimensional space with Python. First, a testing procedure was carried out to validate if the optimizations work as intended through a simple 2D problem. We also showed that the optimal solutions can be improved by using a local coordinate system when the object is far from the global coordinate system. As a more realistic problem, we presented an example of determining the locations of three objects with dimensional errors in 3D space. In addition, the comparisons were discussed in two cases: 1) when determining the positions of the multiple objects at once and 2) when placing the objects one by one in order