Human Intervention Torque Estimation for a Retrofit Steering Actuation System in Autonomous Vehicles

Paper #:
  • 2018-01-0767

  • 2018-04-03
Human steering intervention is an important factor for the driving safety and control performance of autonomous vehicles. Accurate identification of human steering torque will enable human drivers to take over the controls from the autonomous driving system whenever they require. However, due to the fact that there are more than one active torques exerting on the steering wheel for an autonomous driving actuation system, the human torque cannot be detected directly and separately via torque sensors. Therefore, effective estimation though the system dynamics can be taken as an alternative measure to achieve a comparatively accurate quantification of the human steering torque. In this paper, an online estimation strategy of human steering intervention torque for a retrofit steering actuation system of autonomous vehicles is presented. The dynamic model of the steering actuation system is firstly established. The human steering torque estimation algorithm is then devised. To eliminate the usage of angular acceleration signal, an auxiliary variable is introduced to modify the algorithm. The stability condition of the algorithm is analyzed afterwards. As a critical parameter of the estimation algorithm, the influence of the estimation gain upon the convergence rate is discussed. Furthermore, three typical steering cases are designed and simulated to evaluate the performance of the proposed estimator. The simulation results validated the effectiveness of the proposed approach to estimate human steering invention torques. These results showed that the designed estimator could be applied in autonomous driving systems for multiple purposes, such as driver intent inference, Advanced Driver Assistance System, and other applications regarding human-vehicle interaction.
SAE MOBILUS Subscriber? You may already have access.
Attention: This item is not yet published. Pre-Order to be notified, via email, when it becomes available.
Members save up to 36% off list price.
HTML for Linking to Page
Page URL

Related Items

Training / Education
Technical Paper / Journal Article
Technical Paper / Journal Article
Training / Education