Trajectory Control for Self-driving Cars Based on Carla Simulator

Abstract

Automated driving has the potential of saving thousands of lives by removing human error from driving. In order to achieve automated driving, a vehicle needs an adequate set of sensors, which will basically be the eyes and ears of the vehicle. Along with an adequate path following a system that is able to process given data and control the steering and throttle as well as the braking system. An automated vehicle has to detect other road users and should be able to interact with them in a safe way, hence it is important to understand what the vehicle’s sensors should be capable of. Research in the field of automated driving has created promising results in the last few years. Some research groups have shown perception systems which are able to capture even complicated urban scenarios in great detail. Yet, what is often missing are general purpose paths or trajectory planners which are not designed for a specific purpose. This makes most trackers cases sensitive that may only be deployed under specific and close to ideal conditions while the unpredictable nature of surroundings makes this not too ideal of a system if we hope to make our systems fully autonomous and cultivate maximum luxury. Initially we had hoped to be able to implement our controller over a given trajectory and optimize it such that it may be able to generate safe trajectories that may be followed by our system with maximum accuracy over CARLA Simulator. The programming was to be done in python using any available python IDE and was run parallel to the CARLA interface. The results over multiple were compared to show the efficiency of the controller.