| Traditional Electronic Stability Control (ESC) for automobiles is usually accomplished through the use of estimated vehicle dynamics from simplified models. Starting with the conventional two degree-of-freedom vehicle model, one can estimate the vehicle states from the driver steering input. From this estimate, vehicle sideslip angle can be found and this is generally used with a threshold value to initiate a control action. The input/output relationship of the model depends heavily on the accuracy of the parameters used and various means to correct model inaccuracies. Specifically, these models depend on the tire cornering stiffness which is prone to change with age and loading of the tires. Moreover, not all consumers will replace the original equipment (OE) tires with the same ones. Vehicle response is also directly related to coefficient of friction between the tire and road which varies with road and tire conditions. These issues may result in the degradation of the effectiveness of the ESC system. At the very least, they may require extensive tuning of the control algorithms.;This thesis proposes a different method for estimating the instability of a vehicle. It is solely based on measurable vehicle dynamic response characteristics including lateral acceleration, yaw rate, speed, and driver steering input. These signals are appropriately conditioned and evaluated with fuzzy logic to determine the degree of instability present. When the "degree of instability" passes a certain threshold, the appropriate control action is applied to the vehicle in the form of differential yaw braking. Using only the measured response of the vehicle alleviates the problem of degraded performance when vehicle parameters change.;Finally, ten case studies of different vehicles, configurations, environments, driver models, and maneuvers are tested with the same ESC strategy to examine the concept of stability control without estimation. Four very different vehicles ranging from a sports car to a sport utility vehicle (SUV) in multiple configurations including degraded rear tires and different loading conditions are used in evaluating the proposed ESC. These vehicles and configurations are subjected to multiple maneuvers including a double lane change and a fishhook maneuver with tire-to-road conditions such as split mu and low mu to simulate slippery road conditions. The main result of this research is the evolution of a new ESC concept where performance is not based on a vehicle model with set parameters that lose effectiveness in estimating the vehicle dynamic states when the vehicle changes. Instead, the algorithm relies only on the current measurable dynamic states of the vehicle to preserve stability. |
