Neural Control and Synaptic Plasticity for Adaptive Obstacle Avoidance of Autonomous Drones
Drones are used in an increasing number of applications including inspection, environment mapping, and search and rescue operations. During these missions, they might face complex environments with many obstacles, sharp corners, and deadlocks. Thus, an obstacle avoidance strategy that allows them to successfully navigate in such environments is needed. Different obstacle avoidance techniques have been developed. Most of them require complex sensors (like vision or a sensor array) and high computational power. In this study, we propose an alternative approach that uses two simple ultrasonic-based distance sensors and neural control with synaptic plasticity for adaptive obstacle avoidance. The neural control is based on a two-neuron recurrent network. Synaptic plasticity of the network is done by an online correlation-based learning rule with synaptic scaling. By doing so, we can effectively exploit changing neural dynamics in the network to generate different turning angles with short-term memory for a drone. As a result, the drone can fly around and adapt its turning angle for avoiding obstacles in different environments with a varying density of obstacles, narrow corners, and deadlocks. Consequently, it can successfully explore and navigate in the environments without collision. The neural controller was developed and evaluated using a physical simulation environment.
This research was supported partly by Center for BioRobotics (CBR) at the University of Southern Denmark (SDU) and Startup Grant-IST Flagship research of Vidyasirimedhi Institute of Science & Technology (VISTEC).
- 1.Ashour, R., Taha, T., Mohamed, F.: Site inspection drone: a solution for inspecting and regulating construction sites. In: Proceedings of the IEEE 59th International Midwest Symposium on Circuits and Systems, pp. 1–4 (2016)Google Scholar
- 2.Sanfourche, M., Le Saux, B., Plyer, A., Le Besnerais, G.: Environment mapping & interpretation by drone. In: Joint Urban Remote Sensing Event, pp. 1–4 (2015)Google Scholar
- 3.Pobkrut, T., Eamsa-ard, T., Kerdcharoen, T.: Sensor drone for aerial odor mapping for agriculture and security services. In: Proceedings of the 13th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology, pp. 1–5 (2016)Google Scholar
- 4.Mori, T., Scherer, S.: First results in detecting and avoiding frontal obstacles from a monocular camera for micro unmanned aerial vehicles. In: Proceedings of the IEEE International Conference on Robotics and Automation, pp. 1750–1757 (2013)Google Scholar
- 5.Sedaghat-Pisheh, H., Rivera, A.R., Biaz, S., Chapman, R.: Collision avoidance algorithms for unmanned aerial vehicles using computer vision. J. Comput. Sci. Coll. 33, 191–197 (2017)Google Scholar
- 6.Magree, D., Mooney, J.G., Johnson, E.N.: Monocular visual mapping for obstacle avoidance on UAVs. In: Proceedings of the International Conference on Unmanned Aircraft Systems, pp. 471–479 (2013)Google Scholar
- 7.Rohmer, E., Singh, S.P.N., Freese, M.: V-REP: a versatile and scalable robot simulation framework. In: Proceedings of the IEEE International Conference on Intelligent Robots and Systems, pp. 1321–1326 (2013)Google Scholar
- 9.Braitenberg, V.: Vehicles: Experiments in Synthetic Psychology (1986)Google Scholar
- 15.Pasemann, F., Huelse, M., Zahedi, K.: Evolved neurodynamics for robot control. In: European Symposium on Artificial Neural Networks, pp. 439–444 (2003)Google Scholar
- 17.Franceschini, N., Ruffier, F., Serres, J., Viollet, S.: Optic flow based visual guidance: from flying insects to miniature aerial vehicles. INTECH Open Access Publisher (2009)Google Scholar
- 19.Zahedi, K., Pasemann, F.: Adaptive behavior control with self-regulating neurons. In: Lungarella, M., Iida, F., Bongard, J., Pfeifer, R. (eds.) 50 Years of Artificial Intelligence. LNCS (LNAI), vol. 4850, pp. 196–205. Springer, Heidelberg (2007). https://doi.org/10.1007/978-3-540-77296-5_19CrossRefGoogle Scholar