Exhaust gas ignition (EGI)-a new concept for rapid light-off of automotive exhaust catalyst

Increasing pressure on lowering vehicle exhaust emissions to meet stringent California and Federal 1993/1994 TLEV emission standards of 0.125 gpm NMOG, 3.4 gpm CO and 0.4 gpm NOx and future ULEV emission standards of 0.04 gpm NMOG, 1.7 gpm CO and 0.2 gpm NOx has focused specific attention on the cold start characteristics of the vehicle's emission system, especially the catalytic converter. From test data it is evident that the major portion of the total HC and CO emissions occur within the first two minutes of the driving cycle while the catalyst is heating up to operating temperature. The use of an electrically heated catalyst (EHC) has been proposed to alleviate this problem but the cost and weight penalties are high and the durability has yet to be fully demonstrated (1)*. This paper describes a method of reducing the light-off time of the catalytic converter to less than 20 seconds by means of an afterburner. The system uses exhaust gases from the engine calibrated to run rich and additional air injected into the exhaust gas stream to form a combustible mixture. The key feature concerns the method of making this combustible mixture ignitable within 2 seconds from starting the engine when the exhaust gases arriving at the afterburner are cold and essentially non-reacting. © Copyright 1992 Society of Automotive Engineers, Inc.

Modelling nonlinear vehicle dynamics with neural networks

The nonlinear modelling ability of neural networks has been widely recognised as an effective tool to identify and control dynamic systems, with applications including nonlinear vehicle dynamics which this paper focuses on using multi-layer perceptron networks. Existing neural network literature does not detail some of the factors which effect neural network nonlinear modelling ability. This paper investigates into and concludes on required network size, structure and initial weights, considering results for networks of converged weights. The paper also presents an online training method and an error measure representing the network's parallel modelling ability over a range of operating conditions. Copyright © 2010 Inderscience Enterprises Ltd.