|
The purpose of this thesis is to explore the simulation of the test route and the actual road environment of the self-driving car running in a closed test field. In recent years, under the development trend of self-driving cars vigorously promoted by the governments of various countries, the testing regulations and standard systems for self-driving cars However, there is no official version output. It is difficult to compare the self-driving capabilities developed by various manufacturers. At present, the development of self-driving cars is more mature, such as Waymon, Uber, Tesla, Baidu and other companies. However, the self-driving cars being developed by various manufacturers mainly use mileage and time as the actual verification basis, while the actual corresponding road conditions and driving policy are relatively uncomparable. Also, because the traffic conditions and road facilities in various countries are not completely the same, Whether driving can adapt to the traffic conditions in various places is yet to be investigated. As of the end of December 109, the Taiwan Smart Driving Test Laboratory has served a total of 10 self-driving cars in the domestic self-driving team. The test content is mainly based on the scenario planned by the self-driving operation experiment plan. And route to enhance the service energy of Taiwan’s smart driving test laboratory, for example, from 108 to 109, the laboratory plans to set up weather simulation scenarios, increasing from 13 test scenarios to 17, providing rental electronic control platform vehicles for manufacturers to develop components, soft target vehicles that can be moved by remote control, locomotive flatbed equipment, etc. , Through the interlaced combination of test field routes and auxiliary facilities, to create a simulation that is more in line with the domestic traffic environment, and drive the development of the overall domestic self-driving technology and integrate it with international standards. This test method is mainly aimed at different self-driving systems. In a closed test field, it is relatively safe and the situation is reproduced with high integrity considerations, and the vehicle control made by the self-driving decision system is collected for behavior analysis and use. This paper According to the statistical analysis of domestic traffic accidents from 105 to 107 [1], 19 common traffic accidents are generally used as examples of self-driving test scenarios. Based on the situational data of different intersections, signs, etc., the actual closed test field route repositioning by personnel is used. Now test and analyze and compare, and provide self-driving car development and revision guidelines in the future.
|