Parametric Modelling & High-Fidelity Algorithms for Vehicle Weight Estimation for Optimized Concept Vehicle Architecture

The concept definition phase of typical vehicle development focuses on the architecture definition and optimization based on different constraints/requirements. With the focus on Sustainability, the architecture optimization process must include “Light-weighting” as an optimization criterion. With only concept vehicle architecture available, the vehicle weight estimation becomes judgmental & inaccurate. This paper aims to address this deficiency with a new analytical approach for vehicle weight estimation. The new approach for vehicle weight estimation is a “bottom-up” approach using parametric models for each system weight with the inputs being the relevant vehicle specifications driving the system engineering. For size/shape-driven (rather than functional) systems, the models are content-based & segment-based. The parametric models are then iterated for multiple architecture concepts & specifications and the optimum concept (meeting all functional & business constraints) is chosen. The parametric models are based on extensive benchmarking studies and regression analyses. Both single-parameter & multiple-parameter regressions are used for parametric modelling. The parameters selection is carried out such that the relationships have sound technical logic and ensuring the (curve) fits always have regression coefficients greater than 0.8. The proposed approach for weight estimation has been piloted in two new projects & two current products. The weight reduction proposed were quite significant. Also, the vehicle weight estimation is presently carried out post detailing of system concepts - the time lag between vehicle & systems concepts definition leading to significant time lag in vehicle weight estimation. But the proposed approach ensures “zero” time lag as the parametric models provide the vehicle & systems’ weights immediately after the vehicle architecture/specification concept is defined.