What is the software architecture of the autonomous driving system?

1. Overall Architecture

Based on my understanding of the unmanned driving system, the software architecture of the unmanned driving system is layered into the following three layers: perception layer, decision layer and control layer.

2. Perception Layer

The perception layer is mainly realized through various sensors and high-precision maps, including vehicle positioning and object recognition. 

The positioning of the vehicle is mainly based on the integration of information such as LiDar, GPS, inertial sensors, and high-precision maps to obtain the exact position of the vehicle, with a positioning accuracy of even up to the cm level. 

Object recognition is mainly achieved using optical radar and binocular cameras;

3. Decision-making level

The input of the decision layer includes the information of the perception layer, the path planning and the information fed back by the control layer, and the decision instructions are issued by adding learning algorithms. The decision instructions include: following the car, overtaking, accelerating, braking, decelerating, steering, turning around, etc.

4. Control Layer

According to the instructions issued by the decision-making layer, the control layer implements specific control of the vehicle, including: throttle control, brake control, steering wheel control and gear control;