Passing maneuver for autonomous vehicles using a convolutional neural network
Abstract
The development of autonomous vehicles has a lot of challenges that in recent years, both the Academy and the Industry have been trying to solve. Some of them are: The following of the road, passing of vehicles, the parking maneuvers, traffic signs detection, the creation of decision-making systems, etc. One of the most outstanding works is (Bojarski et all., 2016) in which a convolutional neural network called "PilotNet'' is designed. It can take images captured by a camera and generate the lateral control of a vehicle to keep it driving in the center of its lane.
In this work, PilotNet’s network architecture was modified to process the data of a 2D LIDAR sensor mounted on the car and it was trained to carry out the passing maneuver from other vehicles both stationary and in motion, also the lane-keeping. The training data set was generated in the Gazebo simulator, with examples of the passing maneuver performed by both a person and a visual feedback PD controller. PD controller's performance was compared versus neural network and it was found that the network has a behavior more similar to that of a person and is faster in performing the maneuver. In addition, it does not require a decision-making system like the PD controller when changing from one lane to another.
Downloads
References
Ackermann, J., Guldner, J., Sienel, W., Steinhauser, R., and Utkin, V. I. (1995). Linear and nonlinear controller design for robust automatic steering. IEEE Transactions on control systems technology, 3(1):132–143.
Berlin, F. U. (2020). Autominy An autonomous model car for education. https://autominy.github.io/AutoMiny/.
Bojarski, M., Testa, D. D., Dworakowski, D., Firner, B., Flepp, B., Goyal, P., Jackel, L. D., Monfort, M., Muller, U., Zhang, J., Zhang, X., Zhao, J., and Zieba, K. (2016). End to end learning for self-driving cars. CoRR, abs/1604.07316.
Bojarski, M., Yeres, P., Choromanska, A., Choromanski, K., Firner, B., Jackel, L. D., and Muller, U. (2017). Explaining how a deep neural network trained with end-to-end learning steers a car. CoRR, abs/1704.07911.
Carvalho, A., Gao, Y., Gray, A., Tseng, H. E., and Borrelli, F. (2013). Predictive control of an autonomous ground vehicle using an iterative linearization approach. In 16th International IEEE conference on intelligent transportation systems (ITSC 2013), pages 2335–2340. IEEE.
Codevilla, F., M¨uller, M., L´opez, A., Koltun, V., and Dosovitskiy, A. (2018). End-to-end driving via conditional imitation learning. In 2018 IEEE international conference on robotics and automation (ICRA), pages 4693–4700. IEEE.
Hubschneider, C., Bauer, A., Weber, M., and Z¨ollner, J. M. (2017). Adding navigation to the equation: Turning decisions for end-to-end vehicle control. In 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC), pages 1–8. IEEE.
ITAM (2019). ITAM-Robotica Eagle Knights-Wiki. https: //github.com/ITAM-Robotica/Eagle_Knights-Wiki/wiki/ Running-the-Simulation/.
Rajamani, R. (2011). Vehicle dynamics and control. Springer Science & Business Media.
Thrun, S., Montemerlo, M., Dahlkamp, H., Stavens, D., Aron, A., Diebel, J., Fong, P., Gale, J., Halpenny, M., Hoffmann, G., et al. (2006). Stanley: The robot that won the darpa grand challenge. Journal of field Robotics,
(9):661–692.
Wu, T., Luo, A., Huang, R., Cheng, H., and Zhao, Y. (2019). End-to-end driving model for steering control of autonomous vehicles with future spatiotemporal features. In 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 950–955. IEEE.
