Autonomous Vehicles Simulator uses artificial neural network as computational model imitating intelligent vehicle moves. Simulator is an desktop application that was implemented with a C++, SFML and Box2D library. Application allows user to create road model, artificial neural network structure and vehicle model that are later used in a process of optimizing neural network using genetic algorithm. Agent interacts with environment through the vehicle detecting road edges. Except from optimizing neural networks, application allows user to test different vehicle models using different neural networks on the same road which may help in determaning the best agent. Below you can find a short list of instructions of how to set up this project locally.
Visual Studio 2019 (Desktop development with C++ workload is required)
Autonomous Vehicles Simulator project can be easily built locally. If you already have Visual Studio 2019 installed all you have to do is clone the repository and its submodules then open Visual Studio project file. You should be able to compile project based on x64 architecture in Debug and Release mode. Project uses SFML and Box2D libraries that are already present in the repository which means that no further actions are required.
This project is licensed under the terms of the MIT license. Implementation started in June 16 of 2021 as my BSc Thesis written in Polish.
If you have any feedback or issues with the project you can always contact me through my email: [email protected]