A Digital Twin Study: Particle Swarm Optimization of ACC Controller for Follow Acceleration Maneuver

A digital twin (DT) is a technology that allows to create a virtual copy of a real-world system or product. This allows real-world activities to be synchronized, validated and production costs to be reduced. In this study, a DT model is developed to simulate the maneuvers of the vehicle in the virtual environment and optimize the vehicles' control parameters within the scope of Advanced Driver Assistance Systems (ADAS) and Automated Driving (AD) applications. In order to create and validate the DT, we focused on the follow acceleration maneuver (FAM). Firstly, road tests were carried out based on the cases, and vehicle dynamics relevant data were collected with the sensors. The collected data were used to generate scenario files containing static and dynamic environmental information, and movements of road users using ASAM OpenDRIVE® (ODR) and ASAM OpenSCENARIO® (OSC) scenario modelling standards, respectively. The Model. CONNECT™ co-simulation platform (CSP), which enables subsystems containing vehicle model, controller, and scenario information to work together, was used to simulate maneuvers regarding the generated scenarios. While creating the DT, the parameters in the vehicle model were tuned by performing the coastdown maneuver as well as FAM. After all, subsystems were integrated into the AVL Model. CONNECT™ (MC), FAM is performed and controller parameters were optimized using the particle swarm optimization (PSO) algorithm. The main goal of this study is to analyze and validate the DT model, then optimize controller parameters of adaptive cruise control (ACC) with the DT.