Hybrid algorithm for determination of dimensional accuracy of automobile front chassis module using laser measurement data

Measurement data obtained from an inspection system need to be analyzed using appropriate algorithms to estimate dimensional errors and to obtain reliable inspection results. This paper proposes a hybrid algorithm based on computational geometric and simplex search techniques to determine the dimensional errors in manufactured components. The proposed hybrid algorithm works in two stages. First, it establishes a measurement coordinate system using basic computational geometric techniques such as convex hull and Voronoi diagrams. Next, the proposed hybrid algorithm modifies the initial measurement coordinate system by the simplex search technique. This helps to reduce the error between design and measurement coordinate systems. After determining the measurement coordinate system, the dimensional errors in the inspected components are estimated to verify the conformity of manufactured components to the design and tolerance specifications. In the present work, the proposed algorithm has been successfully implemented for inspecting the automobile chassis module. The required measurement data are obtained from a laser vision system developed for inspecting the automobile front chassis module that comprises the subframe and lower arm assembly. The measurement data from the subframe are used for establishing the measurement coordinate system. The dimensional errors in the lower arm are determined using the measurement coordinate system, established earlier. The details of the algorithm and the results obtained for the sample data are presented in this paper. The dimensional error values obtained using the proposed algorithms are compared to those obtained using the least-squares method.