Many vehicles equipped with autonomous driving functions have been developed to increase driver convenience and secure safety, and are being driven on actual roads. Various sensors such as camera, radar and lidar to recognize surrounding objects are applied on the autonomous driving-based vehicles, and advanced technology to control the vehicles using data collected from the sensors is grafted on them. To identify the cause of an accident in autonomous vehicles, it is necessary to compare and analyze a large of scenarios. However, there are constraints of time and cost in the traditional field investigation, and as a result, it causes limitations in defect investigation. This paper will present the necessity of autonomous driving simulation and its methodology for analysis of the autonomous driving accident. The simulation software is Prescan which has been developed by SIEMENS.