Electronic stability control (ESC) is becoming a standard technology in passenger cars, and because of its effectiveness in reducing serious traffic accidents, its use is spreading to commercial vehicles. ESC is activated by comparing the target vehicle behavior, which is analyzed based on driver operation, with the actual vehicle behavior, which is detected by onboard sensors. The target vehicle behavior is analyzed with an initial steer characteristic (such as a vehicle stability factor Ks) representing a vehicle turning characteristic (understeer, neutral steer, or oversteer). Because Ks changes depending on the payload loading states (such as vehicle mass and location of center of gravity), it varies considerably for commercial vehicles.
We analyzed the influence of error in Ks on ESC performance and established a method for the online estimation of Ks. ESC includes the restriction functions of oversteer and understeer. When the preset Ks value differs from the actual value, one of these functions tends to be activated, and activation of the other function is difficult. To compensate for the error in Ks, we developed an algorithm by using a histogram of the estimated values. To improve the estimation accuracy of this estimation algorithm, we found that uncertain estimated values can be eliminated when there is no difference between the actual vehicle behavior and its behavior at the tire friction limit. We verified the effectiveness of this online estimation through vehicle tests.