중력 센서를 이용한 모델 기반 노면 경사도 추정기법

Precise estimation of vehicle mass and road grade are crucial issues in longitudinal control of automotives, such as drivability, traction, or cruise control. There are two main methods for calculating road slope, which is model based road slope estimation and g-sensor based road slope estimation. Vehicle model based road slope estimation needs exact vehicle parameter values and many restriction for slope estimation such as gear shifting, inaccurate estimated engine torque, etc. On the other hand, G-sensor based slope estimation can be calculated most of driving condition but there can be error with real road slope by sensor offset or by vehicle suspension setting. In this paper, a recursive least square method is implemented to the model based road slope estimation for offline parameter estimation and bias adaptation for the g-sensor based road slope estimation will be followed. With bias adaptation, g-sensor based road slope estimation can be reliable for most of driving condition.