Vehicle control utilizing an improved model of tire contact patch kinematics

Vehicle dynamics depends heavily on the tire characteristics. Accordingly, a number of tire models were developed to capture the tire behaviors. Among them, the empirical tire models, generally obtained through lab tests, are commonly used in vehicle dynamics and control analyses. However, these models explain static tire behaviors and overlook the behaviors during transients. They often do not reflect the actual dynamic interactions between tire and vehicle under real operational environments, especially at low vehicle speeds. For example, such models do not predict the vehicle longitudinal overshoots and oscillations immediately after stopping a vehicle with hard braking.;This dissertation proposes a dynamic-deflection tire (DDT) model that not only specifies the known tire characteristics but also captures the dominant transient tire properties in longitudinal and lateral axes. This DDT model can be incorporated with the conventional vehicle models to accurately predict the resonant modes in the vehicle motions, steering lag behavior, and "tire mode switching" behavior. A snowblower has been tested as an example and the data verifies the predictions from the resulting vehicle lateral model. When this snowblower model is utilized in the simulations, the results indicate that the often-ignored tire characteristics can impact the steering control designs for vehicle lane-keeping maneuvers at low speeds. To show the importance of the transient tire dynamics to vehicle longitudinal control, the H-infinity control is employed to design controllers with and without the specific tire dynamics. When applying the two controllers to a passenger car, the experimental results demonstrate that the missing transient tire behavior is a key factor in designing vehicle longitudinal controllers for low-speed applications.;The transient tire behavior also influences passenger ride quality. Ride comfort at low vehicle speed is often overlooked but is important to vehicle control applications, e.g. the latest stop-and-go function in adaptive cruise control. Most control strategies that address passenger comfort simply utilize the bounds of jerk and acceleration of the vehicles. In general, they have several major limitations when applied to low-speed applications: (I) frequency-domain comfort requirements are not integrated and (II) the vehicle models are simplified too far to capture the tire and suspension dynamics that may impact comfort significantly at low speeds. With the aid of the improved vehicle longitudinal model, the impact of the vehicle resonant modes to ride comfort is investigated. A control scheme is developed to address the transient tire behaviors for ride quality under stop-and-go situations. This scheme is based on optimal control and it ensures smooth acceleration during vehicle maneuvers. A two-degree-of-freedom control strategy is used to approximate the optimal control law. Experimental results demonstrate the effectiveness of this control scheme.