Autonomous Vehicles Scenario Testing Framework and Model of Computation: On Generation and Coverage

Autonomous vehicle (AV) technology started to shift the perception of transportation systems. However, for AVs operate at their optimum capabilities, they need go through a comprehensive testing and verification process. While large amount research funding has been provided solving this problem, there is still lack systematic method develop standardized tests that can be used judge if decision-making capability functions are within acceptable parameters. To end, cover all possible situations an AV may run into. This paper focuses on defining notion coverage mathematically when using pseudo-randomly generated simulations testing. The approach defines new equivalence relations between scenes, which systems’ various states, achieve goal. Considering substantial computation, even with obtained coverage, we also introduce mathematical definition sub-scene additional strategies, such as expanding classes scenes combining actors in reduce required certify AVs.