Search Results

A transient, one-dimensional lean NOx trap (LNT) model is described and implemented for vehicle systems simulations. The model accounts for conservation of chemical species and thermal energy, and includes the effects of O₂ storage and NOx storage (in the form of nitrites and nitrates). Nitrites and nitrates are formed by diffusion of NO and NO₂, respectively, into sorbent particles, and reaction rates are controlled by chemical kinetics and solid-phase diffusion. The model also accounts for thermal aging and sulfation by means of empirical correlations, which have been derived from laboratory experiments. Example simulation results using the Powertrain Systems Analysis Toolkit (PSAT) are presented.

Current artificial intelligence techniques for end to end driving of autonomous vehicles typically rely on a single form of learning or training processes along with a corresponding dataset or simulation environment. Relatively speaking, success has been shown for a variety of learning modalities in which it can be shown that the machine can successfully “drive” a vehicle. However, the realm of real-world driving extends significantly beyond the realm of limited test environments for machine training. This creates an enormous gap in capability between these two realms. With their superior neural network structures and learning capabilities, humans can be easily trained within a short period of time to proceed from limited test environments to real world driving.

Lean Gasoline Direct Injection (LGDI) combustion is a promising technical path for achieving significant improvements in fuel efficiency while meeting future emissions requirements. Though Stoichiometric Gasoline Direct Injection (SGDI) technology is commercially available in a few vehicles on the American market, LGDI vehicles are not, but can be found in Europe. Oak Ridge National Laboratory (ORNL) obtained a European BMW 1-series fitted with a 2.01 LGDI engine. The vehicle was instrumented and commissioned on a chassis dynamometer. The engine and after-treatment performance and emissions were characterized over US drive cycles (Federal Test Procedure (FTP), the Highway Fuel Economy Test (HFET), and US06 Supplemental Federal Test Procedure (US06)) and steady state mappings. The vehicle micro hybrid features (engine stop-start and intelligent alternator) were benchmarked as well during the course of that study.

We utilize physically-based models for diesel exhaust catalytic oxidation and urea-based selective catalytic NOx reduction to study their impact on drive cycle performance of hypothetical light-duty diesel-powered hybrid and plug-in hybrid vehicles (HEVs and PHEVs). The models have been implemented as highly flexible SIMULINK block modules that can be used to study multiple engine-aftertreatment system configurations. The parameters of the NOx reduction model have been adjusted to reflect the characteristics of commercially available Cu-zeolite catalysts, which are of widespread current interest. We demonstrate application of these models using the Powertrain System Analysis Toolkit (PSAT) software for vehicle simulations, along with a previously published methodology that accounts for emissions and temperature transients in the engine exhaust.

We summarize results from comparative simulations of hybrid electric vehicles with either stoichiometric gasoline or diesel engines. Our simulations utilize previously published models of transient engine-out emissions and models of aftertreatment devices for both stoichiometric and lean exhaust. Fuel consumption and emissions were estimated for comparable gasoline and diesel light-duty hybrid electric vehicles operating over single and multiple urban drive cycles. Comparisons between the gasoline and diesel vehicle fuel consumptions and emissions were used to identify potential advantages and technical barriers for diesel hybrids.

Driver is a key component in vehicle simulation. An ideal driver model simulates driving patterns a human driver may perform to negotiate road profiles. There are simulation packages having the capability to simulate driver behavior. However, it is rarely documented how they work with road profiles. This paper proposes a new truck driver model for vehicle simulation to imitate actual driving behavior in negotiating road grade and curvature. The proposed model is developed based upon Gipps' car-following model. Road grade and curvature were not considered in the original Gipps' model although it is based directly on driver behavior and expectancy for vehicles in a stream of traffic. New parameters are introduced to capture drivers' choice of desired speeds that they intend to use in order to negotiating road grade and curvature simultaneously. With the new parameters, the proposed model can emulate behaviors like uphill preparation for different truck drivers.

Traditional autonomous vehicle perception subsystems that use onboard sensors have the drawbacks of high computational load and data duplication. Infrastructure-based sensors, which can provide high quality information without the computational burden and data duplication, are an alternative to traditional autonomous vehicle perception subsystems. However, these technologies are still in the early stages of development and have not been extensively evaluated for lane detection system performance. Therefore, there is a lack of quantitative data on their performance relative to traditional perception methods, especially during hazardous scenarios, such as lane line occlusion, sensor failure, and environmental obstructions.