{"id":5849,"date":"2021-06-22T18:13:48","date_gmt":"2021-06-22T16:13:48","guid":{"rendered":"https:\/\/blogs.mathworks.com\/student-lounge\/?p=5849"},"modified":"2021-06-22T18:13:48","modified_gmt":"2021-06-22T16:13:48","slug":"driving-an-autonomous-vehicle-on-a-track-with-recorded-synthetic-data","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/student-lounge\/2021\/06\/22\/driving-an-autonomous-vehicle-on-a-track-with-recorded-synthetic-data\/","title":{"rendered":"Driving an Autonomous Vehicle on a Track with Recorded Synthetic Data"},"content":{"rendered":"

Today\u2019s blog post is written by Veer Alakshendra, Education Technical Evangelist on the Student Competition team at MathWorks. <\/em><\/p>\n

With\u00a0advancements in the automotive industry, various student competitions have introduced\u00a0the\u00a0driverless category,\u00a0where the goal of the teams is to\u00a0design and build an autonomous vehicle that can compete\u00a0in different disciplines.\u00a0<\/span>\u00a0<\/span><\/p>\n

Using\u00a0<\/span>MATLAB and Simulink<\/span><\/a>,<\/span>\u00a0you can design automated driving system functionality including sensing, path planning, sensor fusion,\u00a0and control\u00a0systems.\u00a0<\/span>\u00a0<\/span><\/p>\n

In this article, we\u00a0will demonstrate an approach\u00a0to drive\u00a0an autonomous vehicle\u00a0in a closed-loop\u00a0circuit.\u00a0The task here is to drive the car in an unknown environment avoiding collision with the cones ensuring to complete the necessary laps.\u00a0<\/span>\u00a0<\/span><\/p>\n

\"\"<\/p>\n

Scene Creation<\/span>\u00a0<\/span><\/h1>\n

The first step is to create\u00a0a 3D simulation environment consisting\u00a0of a vehicle,\u00a0track, and cones.\u00a0The\u00a0<\/span>Vehicle Dynamics Blockset<\/span><\/a>\u00a0Toolbox\u00a0comes installed with prebuilt 3D scenes to simulate and visualize the vehicles\u00a0modeled in Simulink. These 3D scenes are visualized using the\u00a0<\/span>Unreal Engine<\/span><\/a>\u00a0from Epic Games.<\/span>\u00a0<\/span><\/p>\n

As the\u00a0current problem requires\u00a0a\u00a0customized scene, we\u00a0used\u00a0the Unreal Editor and\u00a0the\u00a0<\/span>Vehicle Dynamics Blockset Interface for Unreal Engine 4 Projects support package<\/span><\/a>\u00a0to build the scene.\u00a0<\/span>\u00a0<\/span><\/p>\n

To learn how to customize scenes\u00a0(Figure 2), please follow the steps explained in the\u00a0<\/span>documentation<\/span><\/a>.\u00a0Alternatively,\u00a0we\u00a0can also use\u00a0<\/span>RoadRunner<\/span><\/a>\u00a0to design 3D scenes for automated driving simulations.\u00a0<\/span>\u00a0<\/span><\/p>\n

\"\"<\/p>\n

Fig 1. Custom scene in Unreal<\/span><\/span><\/em>\u00a0<\/span><\/p>\n

\"\"<\/p>\n

Fig 2.\u00a0<\/span>Steps for creating\u00a0<\/span>a\u00a0<\/span>custom scene<\/span><\/span><\/em>\u00a0<\/span><\/p>\n

Lap1: Environment Mapping<\/span><\/span>\u00a0<\/span><\/h1>\n

\"\"<\/p>\n

Fig 3. Simulink model for\u00a0environment mapping\u00a0<\/em><\/p>\n

The\u00a0next task\u00a0is to map the environment. As mentioned in the previous section, the\u00a0driverless vehicle is in an\u00a0unknown\u00a0environment\u00a0that\u00a0consists of cones kept on both sides of the track.\u00a0To\u00a0detect the\u00a0cones and generate a reference path for the\u00a0first\u00a0lap, we have built\u00a0a Simulink model as shown in\u00a0Figure\u00a03.\u00a0Figure\u00a04\u00a0shows the steps\u00a0performed by the model\u00a0in the first lap:\u00a0<\/span>\u00a0<\/span><\/p>\n

\"\"<\/p>\n

Fig 4.\u00a0Block diagram representation of environment mapping\u00a0\u00a0<\/em><\/p>\n