Hierarchical Reinforcement Learning Approach for the Road Intersection Task

The task of building unmanned automated vehicle (UAV) control systems is developing in the direction of complication of options for interaction of UAV with the environment and approaching real life situations. A new concept of so called “smart city” was proposed and view of transportation shifted in direction to self-driving cars. In this work we developed a solution to car’s movement on road intersection. For that we made a new environment to simulate a process and applied a hierarchical reinforcement learning method to get a required behaviour from a car. Created environment could be then used as a benchmark for future algorithms on this task.