Autonomous driving simulation

NegotiableUpdate on 05/07

Model

Nature of the Manufacturer: Producers

Product Category

Place of Origin

Online Consult

Overview

The ANSYS high-precision autonomous driving simulation verification platform provides physics based 3D scene modeling, semantic based road event modeling, simulation of cameras and LiDAR based on physical optical properties, and simulation of millimeter wave radar based on physical electromagnetic properties, thereby achieving real-time closed-loop simulation of multiple sensors, multiple traffic objects, multiple scenes, and multiple environments.

Product Details

System solution

ANSYS high-precisionAutonomous driving simulationThe verification platform provides physics based 3D scene modeling, semantic based road event modeling, simulation of cameras and LiDAR based on physical optical properties, and simulation of millimeter wave radar based on physical electromagnetic properties, thereby achieving real-time closed-loop simulation of multiple sensors, multiple traffic objects, multiple scenes, and multiple environments. Its main functions are as follows:

1. Open traffic scene editing module, customizable road and traffic scenes, can customize buildings, green belts, and so on on on both sides of the road;

2. According to user needs, the traffic flow on the road scene can be customized, and the vehicles, pedestrians, and traffic lights coming and going on the road can be customized;

3. Vehicle dynamics parameters for active driving (or algorithm controlled vehicles) can be set according to customer needs;

4. Support high-precision 3D scene simulation and simulation based on ambient light;

5. Support high-precision sensor simulation of physical attributes, including simulation of millimeter wave radar, camera, and lidar;

In addition, considering the ability to more realistically reflect the simulation testing of "human vehicle road" in the environment, the platform also provides an open interface that can seamlessly integrate with physical sensors, VR devices, controllers, and various testing data, thus better meeting the testing and simulation requirements of different levels and goals.

System Composition

Below are the components of each module on this platform.

Custom Road Environment ANSYSAutonomous driving simulationThe platform provides a set of design tools for custom road scenes, with the ability to design straight roads, curves, and curves. It supports editing of road width, length, radius, direction, number of lanes, lane direction, lane speed limit, lane type, and more.

At the same time, the design tool supports the definition of elevated and other roads of different heights, as well as different slope angles, road intersections, ramps, and merging roads. It also supports custom modeling of lane lines, including single line, double line, solid line, dashed line, lane line texture, color, and a series of lane line types. At the same time, the software integrates a rich library of environmental models, such as trees, buildings, traffic signs, streetlights, utility poles, green belts, animals, obstacles and facilities on construction roads, traffic pedestrians, and other object models, which can quickly model road scenes according to user needs.

In addition to custom scenes, the ANSYS platform also supports importing 3D high-precision maps such as OpenStreetMap to automatically generate road models that match the maps.

Customize traffic scenarios

The ANSYS platform also provides fast semantic based road traffic flow design, including lane driving rules, vehicle and pedestrian behavior, traffic sign behavior, and accurate data output of traffic behavior of various traffic objects at a certain moment. In addition, the behavior of traffic objects can also be artificially defined, including simulation of a series of scenarios such as vehicle driving behavior, sudden lane changes, sudden acceleration, pedestrians running red lights indiscriminately, and sidewalks. At the same time, the interaction between vehicles and pedestrians within the software can be customized to simulate behaviors such as automatically avoiding pedestrians and ignoring pedestrian collisions. The software is embedded with scripting language definitions and also supports interface control in languages such as Python and C++to define traffic behavior. As shown in the figure below, the behavior of traffic objects such as vehicles and pedestrians is defined through a semantic level scripting language.