Apollo automatic driving advanced learning: how to realize obstacle bypass and parameter debugging

First of all, you need to understand that the Planning module is based on the levels of Scenario, Stage, and Task, namely: Scenario -> Step -> Specific decision-making method. Apollo can cope with different road scenarios faced by autonomous driving, all of which are registered and managed through Scenario. Scenario uses a finite state machine to judge and select the current driving scene. There are multiple Stages under each Scenario, which refer to the rough steps that need to be executed in the current scene.

The Planning module plans a trajectory for the self-driving vehicle based on the surrounding environment information perceived by routing (navigation module), prediction (perception module), and map positioning and navigation information (including coordinates, speed, acceleration, jerk, time, etc.) , and then pass this information to the control module.

1. Requirements

During the driving process of the main vehicle, other types of obstacles are detected in the front and then take a detour. The horizontal distance from the target obstacle should be kept at least 1 meter, and the speed limit for avoiding obstacles should not exceed 5m/s.

2. Analysis

2.1 Part of the code related to the lateral distance of overtaking