Development of autonomous vehicle control algorithm based on DGPS(RTK) and test vehicle performance verification

Recently, advanced automobile parts with functions of ASV (Advanced Safety Vehicle) and ADAS (Advanced Driving Assistance System) are developed by various automotive manufactures and major parts companies. To develop and commercialize these advanced parts, various test methods including unit function test, compatibility test, VDS (Vehicle Driving Simulator) based test and T-Car test are needed because reliability is the most important factor for safety. But it is very difficult to make same experimental condition in T-car test as a final test, because driver feels hardness to repeat movement of the vehicle such as speed and path of travel. So we develop the driving robot adaptable to various kinds of vehicles, speed control and path following algorithm based on RTK-DGPS(Real Time Kinematic-Differential Global Positioning System) to guarantee repeatability of test condition in T-car test. The proposed speed control algorithm is used to make the vehicle keep its velocity and decrease overshoot of vehicle speed. The proposed path following algorithm makes the vehicle pass along the predefined path. Two algorithms don't need to any additional sensor except RTK-DGPS and the vehicle dynamic model. So it is very convenient to apply these algorithms to various cars with different dynamic characteristics. Developed algorithms are verified by a computer simulation and the real-car test.