Abstract
Lateral velocity is an important parameter to characterize vehicle stability. The acquisition of lateral velocity is of great significance to vehicle stability control and the trajectory following control of autonomous vehicles. Aiming to resolve the problems of poor estimation accuracy caused by the insufficient modeling of traditional model-based methods and significant decline in performance in the case of a change in road friction coefficient, a deep learning method for lateral velocity estimation using an LSTM, long-term and short-term memory network, is designed. LSTM can well reflect the inertial characteristics of vehicles. The training data set contains sensor data under various working conditions and roads. The simulation results show that the prediction model has high accuracy in general and robustness to the change of road friction coefficient.
Highlights
With the continuous growth of car ownership, the frequency of traffic accidents has increased year by year [1]
The performance of the estimator was evaluated under four friction coefficients and three vehicle speeds according to the double line change (DLC) condition of ISO3888-1 standard
With the de decrease of the road friction coefficient, theaccurately deviation ofestimate the estimated value will not change crease of the road friction coefficient, the deviation of the estimated value will not chang significantly, and it is robust to the change of the road friction coefficient
Summary
With the continuous growth of car ownership, the frequency of traffic accidents has increased year by year [1]. To improve vehicle safety and reduce the frequency of accidents, a large number of active control systems (ACSs) are assembled in vehicles during mass production. These systems mainly include active front wheel steering (AFS), electronic stability program (ESP), and traction control systems (TCS). With the continuous improvement of vehicle intelligence, advanced driver assistance systems (ADAS) are widely used [2], including adaptive cruise control (ACC) and lane keeping assistance (LKA) systems. The advanced active control systems and advanced driver assistance systems are achieved based on the acquisition of some basic states of the vehicle [1,3], such as vehicle sideslip angle, yaw rate, longitudinal speed, and lateral speed, etc. Due to limited sensor accuracy and cost as well as the difficulty in determining the distribution characteristics of measurement noise, some states or parameters cannot be effectively measured using applicable sensors, or measurement performance cannot meet the accuracy requirements
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
