Search Results

Vehicle Rollover Prevention/Warning Systems have recently been an important topic in Advanced Driver Assistance Systems (ADAS) of automotive electronics field. This paper will propose a rollover-prevention system implementation with vehicle dynamic model, video-detection technique and rollover index to help the driver avoid accidents as driving into a curve. Due to the reason that vehicle rollover motion analysis needs complicated computation and accurate parameters of vehicle stability in real time, in the first stage a vehicle dynamic model based on Extended Kalman Filter (EKF) algorithm is built, which can estimate vehicle roll/yaw motion in the curve by vehicle sensors. And then the image-based technique will be employed in detecting the front road curvature, and combined in the system to predict vehicle steering status. The final stage is to apply the vehicle rollover index with estimated vehicle motion to predict the dangerous level to drivers for warning.

In recent years, many attentions have been paid on global environmental protection and energy saving; more people, therefore, have chosen bikes for commuting to work or school. For longer distance transportation and less effort, electric power assist bikes have re-entered the market. Due to regulation of some countries, electric bikes that must be pedaled were developed. These machines utilize the pedals as the dominant form of propulsion, with the motor used only to give extra assistance when needed for hills or long journeys. The ratio of electric power to human power may affect the riding feel. As a result, a torque sensor, which detects the pedaling force, is crucial in this application. This paper proposes a new design of torque sensor by way of twist angle measurement. It is composed of a torsion bar, input and output shaft with ring magnets and Hall sensors to achieve contactless sensing.