Automatic Driving Control Method for Lane Departure in Expressway Ramp Merge Area

The traffic volume in the merging area of the expressway ramp is large, which leads to the perception difficulty of the auto drive system. Therefore, an automatic driving control method for lane deviation in the merging area of expressway ramps is proposed. Comprehensively perceive the surrounding environment through multiple sensors and collect vehicle driving and surrounding environment data. Perform two-step region of interest extraction and median filtering preprocessing on road images to reduce noise and interference information. Use the Sobel edge detection operator and Hough transform method to detect lane markings and obtain parameters. Establish a lane departure warning model and adopt a dual warning strategy to achieve accurate warning. Based on model predictive control strategy, the target steering torque is determined to achieve lane departure autonomous driving control. The experimental results show that the accuracy of lane departure warning using this method is over 95%, and the standard deviation of vehicle lane departure is below 0.05[Formula: see text]m, which improves the safety and stability of autonomous driving in complex environments.