Development of a Wheel Force Transducer force data prediction algorithm using vehicle driving image data

Wheel Force Transducer is essential in the vehicle research and development process, such as the design and reliability evaluation of the vehicle chassis and related parts. Force values obtained from WFT depend on the various driving environments and roads in which the vehicle travels, and measuring this data requires a lot of cost and manpower.This paper deals with the research content that predicts the force data of WFT based on vehicle driving image data. To extract temporal and spatial features from driving image data at once, we propose a deep learning model-based algorithm using 3D CNN with high performance in the field of video data-based behavior recognition. By conducting this study, we intend to contribute to more efficient vehicle research and development by reducing the cost and manpower required for vehicle development.