Data-driven Customisable Control Design for an Adaptive Cruise Control System

This study proposes a data-driven controller design for Adaptive Cruise Control (ACC) systems. The design employs Long Short-Term Memory (LSTM) networks to enable dynamic customisation and superior performance over classical controllers like PID. The proposed LSTM-based controller leverages driving data to optimise control actions, allowing for adaptive and flexible responses in various traffic conditions. The training data for the LSTM network includes the data captured from a desktop motion simulator integrated with MATLAB's vehicle dynamic block set simulating the motion of the ego vehicle. The results show that the LSTM controller effectively maintains safe following distances and smooth velocity transitions, outperforming traditional methods in terms of adaptability and response time. The LSTM-based controller consistently delivers more stable control actions, particularly in scenarios involving significant velocity variations. These findings highlight the potential of data-driven approaches in enhancing the customisation, efficiency, and safety of ACC systems in real-world applications.