Automated Robustness Simulation Testing of an Autonomous Vehicle

Abstract

Autonomous vehicles are facing a significant problem when it comes to testing different types of scenarios with various parameters, e.g. road friction and road slopes. It is crucial that autonomous vehicles can handle these scenarios to assure robustness of the system. In this study, this problem is addressed by developing a simulation environment for an autonomous vehicle model and testing the robustness of the model by applying one fault, steering miss alignment, and one condition, various road surfaces. Two algorithms, namely Search Based testing and Monte Carlo are involved in manipulating the values of these parameters, to find the best combination of parameters that gave the highest deviation values. The algorithms are later compared on how well they test the robustness of the autonomous vehicle model by comparing these deviation values.