The behavioral decision-making task aims to determine the high-level actions that the autonomous vehicle should take based on the current environmental perception information and driving goals, such as acceleration, deceleration, lane change, parking, etc. The difficulty of this task lies in the need to comprehensively consider traffic rules, road conditions, the behavior of traffic participants, and the status of the vehicle itself to make a safe and reasonable decision. Inputs usually include environmental perception information, prediction results of other traffic participants, and status information of the own vehicle (such as speed, location, etc.). The output is a series of high-level action instructions to guide the vehicle on how to respond to current traffic conditions.

The purpose of the motion planning task is to calculate the acceleration and turning angle information that the vehicle should take based on the complete trajectory containing spatial coordinates and time information. The difficulty of this task lies in the need to consider the dynamic characteristics of the vehicle while satisfying the path constraints to ensure that the trajectory is smooth and the control actions are executable. Inputs typically include the path information output by path planning, the vehicle's dynamics model, and any real-time constraints (such as speed limits, acceleration limits, etc.). The output is the acceleration and angle the vehicle should take.