Design and Implementation of a Reconfigurable Virtual Hybrid Electric Vehicle Simulator, 03-9196Printer Friendly Version
Inclusive Dates: 07/01/00 - 09/30/00
Background - World competition and stringent United States fuel economy goals and emission regulations for the 21st Century vehicle have pressured the automotive industry to design and evaluate advanced automobiles at an accelerated rate. The industry consensus is that the hybrid electric vehicle is the only technology that will meet near-zero emission and 80-miles/gallon fuel economy goals. However, hybrid electric vehicles (HEV) operate much differently than conventional vehicles. Therefore, existing design techniques and guidelines developed for conventional powertrains do not apply well to hybrid vehicles. Numerous sources in the literature support the conclusion that rapid development of these vehicles has been hampered by a number of factors, including:
The tedious and time-consuming tasks required to overcome the deficiencies of current HEV development approaches and the numerous failed HEV development programs are convincing evidence that the automotive industry is in need of a reconfigurable and modular tool to design and accurately evaluate hybrid systems. Currently, no commercially available tool for prototyping HEV systems offers the flexibility, or even the power and computing capability, to simultaneously assess and optimize the performance of multiple HEV subsystems.
Approach - The research team has conceived a virtual HEV simulator concept that uniquely combines vehicle analysis, design, and testing on a single environment. The simulator provides a link between high-fidelity models (software) and two types of physical systems (hardware): a high-power mechanical unit (>100 kW) and a high-power electrical unit (up to 530 Vdc and 420 Adc). This approach creates a rapid and cost-effective interface system directly targeted to significantly expedite HEV design evaluations and control system optimization. The Reconfigurable Virtual HEV Simulator will not only expedite hybrid vehicle component sizing through rigorous scaling algorithms, but also provide a flexible means for optimizing and hardware-testing critical control units individually and collectively. Furthermore, this multiple-microprocessor network eventually will be able to communicate to the original manufacturer's engine, motor, and transmission control modules via different communication buses.
Accomplishments - The first phase of the program has been to develop a set of scaling algorithms that expedites the evaluation of different hybrid configurations. An automated algorithm has been developed to reduce laborious user interface. Additionally, a parallel stochastic optimization scheme is currently being developed in an engineering software utility that minimizes fuel consumption and maximizes performance. User input is virtually reduced to a handful of component sizes (actually only a range of sizes is needed), and the rest of the design process is basically accomplished on its own.