Date of Award

12-2016

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Automotive Engineering

Committee Member

Bashah Ayalew, Committee Chair

Committee Member

Yunyi Jia

Committee Member

Robert Prucka

Committee Member

Paul Venhovens

Abstract

The objective of this research is to develop an integrated system engineering methodology for the customization design to maximize vehicle performance upgrade freedom while ensuring vehicle dynamics compliance. A post-delivery modification framework, which is led by an aftermarket umbrella organization and involve various stakeholders has been established. The umbrella organization will be in charge of developing the design envelope and distribute to various aftermarket kit suppliers to generate specific products according to their brand essence. A generic mathematical representation of a (proprietary) ESC system has been developed for virtual certification purposes. This approach is a cost-effective alternative to physical on-road testing and hardware-in-the-loop (HiL) simulations. Furthermore, based on the stability control model, the modification impacts on the vehicle dynamics and stability performance was assessed using the Taguchi design of experiment (DOE) method. DOE results provide three distinct ways for supporting aftermarket modifications. First, main effects help customizers to understand which modification bring benefits or risks. Second, a regression model of the lateral offset metrics helps suppliers to predict closed-loop performances with open-loop testing information which require much less time and cost. Finally, the pass/fail criteria regarding federally mandated ESC compliance (FMVSS 126) brought on the ‘Pass Region' which consisted of feasible configurations such that customizers may configure their options within a safe zone. Each of these methods complements others for supporting the aftermarket modification. In order to improve the computation efficiency, two lower fidelity models were developed: A linear model and a surrogate model. The linear model is derived from the high fidelity model with reduced degrees of freedom (DOF) and linearized parameters. Tire cornering stiffness is treated as constants for gentle maneuvers, and varying parameters for high-dynamic driving maneuver. The linear system is either a linear time-invariant (LTI) system or a linear parameter-varying (LPV) system depending on the application context. The PD yaw stability control algorithm, which is inherited from the high fidelity model, was simplified but retained with critical nonlinear features. A quadratic regression model that was dedicated for compliance metrics was developed as a surrogate model incorporated in an interactive rule-based design envelope. An interactive design envelope has been created incorporating the rules established using computational efficient linear and surrogate models. The constraint satisfaction problem is described in the nonlinear programming context and solved using sequential quadratic programming. The quasiconvexity of the design space, which is the necessary condition for the proposed approach, is also investigated by inspecting the constraint functions. Finally, two case studies were developed to demonstrate the framework developed which was validated against the high fidelity co-simulation model.

Share

COinS