Getting started with self-driving planning

This chapter focuses on the planning layer of autonomous vehicle systems, which consists of three layers: task planning, behavior planning, and motion planning. Task planning is also called path planning. In task planning, the A∗ algorithm, as a discrete space search algorithm, is primarily used to solve the problem of optimal path search in discrete space. Behavior planning is a core part of behavior planning for autonomous vehicle systems. The finite state machine (FSM) and hierarchical FSM are often used to design a behavioral decision model. To make the motion planning easy to understand, we illustrate a series of methods for spline interpolation.