Data-Driven Dimensional Synthesis of Diverse Planar Four-bar Function Generation Mechanisms via Direct Parameterization

2025-07-10

AI Summary

Planar four-bar linkage dimensional synthesis is a classic inverse kinematics problem requiring the inference of mechanism parameters from prescribed trajectory constraints. This paper proposes a data-driven, end-to-end supervised learning framework that integrates Long Short-Term Memory (LSTM) networks with a type-aware Mixture-of-Experts (MoE) architecture, enabling unified modeling and generation across multiple linkage types—including crank-rocker, double-crank, and double-rocker mechanisms. We introduce a novel motion-simulation-based metric and combine synthetic-data training with kinematic feedback to refine generation quality. Experiments demonstrate that our method efficiently produces high-accuracy linkage parameters free from assembly or mobility defects. It significantly lowers the design barrier for non-expert users while improving both synthesis efficiency and practical feasibility.

Technology Category

Intelligent Robots: Motion and Path PlanningMachine Learning: Mixture of Experts (MoE)Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSemantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsGraph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Dimensional synthesis of planar four-bar mechanisms is a challenging inverse problem in kinematics, requiring the determination of mechanism dimensions from desired motion specifications. We propose a data-driven framework that bypasses traditional equation-solving and optimization by leveraging supervised learning. Our method combines a synthetic dataset, an LSTM-based neural network for handling sequential precision points, and a Mixture of Experts (MoE) architecture tailored to different linkage types. Each expert model is trained on type-specific data and guided by a type-specifying layer, enabling both single-type and multi-type synthesis. A novel simulation metric evaluates prediction quality by comparing desired and generated motions. Experiments show our approach produces accurate, defect-free linkages across various configurations. This enables intuitive and efficient mechanism design, even for non-expert users, and opens new possibilities for scalable and flexible synthesis in kinematic design.

Problem

• Solves inverse kinematics for four-bar mechanisms using data-driven methods
• Determines mechanism dimensions from desired motion specifications via learning
• Enables accurate defect-free linkage design for diverse configurations

Innovation

• LSTM-based neural network for sequential precision points
• Mixture of Experts architecture for linkage types
• Novel simulation metric for motion comparison