Search Results

This paper proposes an approach that characterizes a driver's driving behavior and style in real-time during car-following drives. It uses an online learning of the evolving Takagi-Sugeno fuzzy model combined with the Markov model. The inputs fed into the proposed algorithm are from the measured signals of on-board sensors equipped with current vehicles, including the relative distance sensors for Adaptive Cruise Control feature and the accelerometer for Electronic Stability Control feature. The approach is verified using data collected using a test vehicle from several car-following test trips. The effectiveness of the proposed approach has been shown in the paper.

This paper proposes an approach to use ABS wheel speed sensor signals together with other vehicle state information from a brake control module to detect an unbalanced tire or tires in real-time. The proposed approach consists of two-stage algorithms that mix a qualitative method using band-pass filtering with a quantitative parameter identification using conditional least squares. This two-stage approach can improve the robustness of tire imbalance or imbalances. The proposed approach is verified through vehicle testing and the test results show the effectiveness of the approach.

Increasing electronics content, growing computing power, and proliferation of opportunities for information connectivity (through improved sensors, GPS, road and traffic information systems, wireless internet access, vehicle-to-vehicle communication, etc.) are technology trends which can significantly transform and impact future automotive vehicle's control and diagnostic strategies. One aspect of the increasing vehicle connectivity is access to ambient and road condition information, such as ambient temperature, ambient pressure, humidity, % cloudiness, visibility, cloud ceiling, precipitation, rain droplet size, wind speed, and wind direction based on wireless internet access. The paper discusses the potential opportunities made available through wireless communication between the vehicle and the internet.

Government regulations for fuel economy and emission standards have driven the development of technologies that improve engine performance and efficiency. These technologies are enabled by an increased number of actuators and increasingly sophisticated control algorithms. As a consequence, engine control calibration time, which entails sweeping all actuators at each speed-load point to determine the actuator combination that meets constraints and delivers ideal performance, has increased significantly. In this work we present two adaptive optimization methods, both based on an indirect adaptive control framework, which improve calibration efficiency by searching for the optimal process inputs without visiting all input combinations explicitly. The difference between the methods is implementation of the algorithm in steady-state vs dynamic operating conditions.