@inproceedings{fb0733246f314782a34a3fd28672f45d,
title = "A GENERAL SIMULATION FRAMEWORK AND PATH SYNTHESIS OF SPATIAL FOUR-BAR MECHANISMS USING DEEP GENERATIVE MODELS",
abstract = "We present an extensible framework for the kinematic synthesis of spatial four-bar mechanisms, integrating a general simulator for dataset generation and a deep generative model for path synthesis. Recent advancements in machine learning have made the unified kinematic synthesis of one-degree-of-freedom (DOF) linkage mechanisms increasingly attractive. However, little research has explored their application to general spatial linkage mechanisms. There are two key challenges in adopting machine learning to spatial linkage mechanisms: the need for a simulator capable of efficiently generating large-scale spatial linkage mechanism datasets, and the development of a deep generative model suited to spatial mechanism synthesis. This paper aims to address both challenges by proposing a framework that combines simulation and machine learning to advance spatial mechanism synthesis.",
keywords = "Deep Generative Model, Linkage Mechanism, Machine Learning, Mechanism Simulation, Path Synthesis, Spatial Mechanism",
author = "Xueting Deng and Anurag Purwar",
note = "Publisher Copyright: Copyright {\textcopyright} 2025 by ASME.; ASME 2025 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2025 ; Conference date: 17-08-2025 Through 20-08-2025",
year = "2025",
doi = "10.1115/DETC2025-168963",
language = "English",
series = "Proceedings of the ASME Design Engineering Technical Conference",
publisher = "American Society of Mechanical Engineers (ASME)",
booktitle = "21st IEEE/ASME International Conference on Mechatronic and Embedded Systems and Applications (MESA); 49th Mechanisms and Robotics Conference (MR)",
}