Vehicle Traction Control Research Articles

Modeling and Validation of Automotive Engines for Control Algorithm Development

There is considerable interest in coordinated automotive engine/transmission control to smooth shifts, and for traction control of front wheel vehicles. This paper outlines a nonlinear dynamic engine model of a port fuel-injected engine, which can be used for control algorithm development. This engine model predicts the mean engine brake torque as a function of the engine controls (i.e., throttle angle, spark advance, fuel flow rate, and exhaust gas recirculation (E. G. R.) flow rate). The model has been experimentally validated for a specific engine, and includes: • intake manifold dynamics, • fuel delivery dynamics, and • process delays inherent in the four-stroke engine. This model is used in real time within a control algorithm, and for system simulation. Also, it is flexible enough to represent a family of spark ignition automotive engines, given some test and/or simulation data for setting parameters.

Stability enhancement and traction control of road vehicles

An algorithm that enhances the stability and traction of a road vehicle (which may be steering) during ‘acceleration’ (meaning either braking or accelerating) is presented. This algorithm has two domains: partial-effort acceleration and maximum-effort acceleration. During partial-effort acceleration, stability enhancement requires only proper distribution of longitudinal traction forces. During maximum-effort acceleration, stability enhancement requires distribution and reduction of longitudinal traction forces. The algorithm attempts to provide the desired levels of acceleration while it simultaneously constrains trajectory disturbance of the vehicle. The algorithm uses total information about the road surface characteristics, the driver's actions, and the vehicle's design and dynamic states. It fully accommodates non-linear vehicle dynamics and tyre force characteristics. The algorithm has the form of a linear program and its operation is traced in a series of graphs where the geometry of linear programming is used as a setting.