A Study on the AI Model for Predicting the Natural Frequency of Knuckle System Using Image

Recently, studies have been actively conducted to predict the performance of systems using artificial intelligence (AI) techniques or to classify/detect systems through image-based data. In this study, we propose an AI model that predicts the natural frequency of the knuckle system which is in vehicle using images and specifications. Natural frequency is a critical aspect of NVH performance, and AI model can predict this performance without the need for an FE model. This allows designers to quickly assess the knuckles’ performance in the early stages of development This AI model utilizes features from knuckle images as learning data to improve prediction accuracy. As a result, the AI-based natural frequency prediction model can significantly increase efficiency in the early stages of vehicle development.