DisGNet: A Distance Graph Neural Network for Forward Kinematics Learning of Gough-Stewart Platform
This addresses a specific robotics problem (Gough-Stewart platform kinematics) with incremental improvements in precision and speed for practical deployment.
The paper tackles the forward kinematics problem of the Gough-Stewart platform by proposing DisGNet, a graph neural network that learns graph distance matrices, combined with a GPU-friendly Newton-Raphson method for refinement. The results show error accuracies below 1mm and 1deg at 79.8% and 98.2% respectively, meeting real-time requirements.
In this paper, we propose a graph neural network, DisGNet, for learning the graph distance matrix to address the forward kinematics problem of the Gough-Stewart platform. DisGNet employs the k-FWL algorithm for message-passing, providing high expressiveness with a small parameter count, making it suitable for practical deployment. Additionally, we introduce the GPU-friendly Newton-Raphson method, an efficient parallelized optimization method executed on the GPU to refine DisGNet's output poses, achieving ultra-high-precision pose. This novel two-stage approach delivers ultra-high precision output while meeting real-time requirements. Our results indicate that on our dataset, DisGNet can achieves error accuracys below 1mm and 1deg at 79.8\% and 98.2\%, respectively. As executed on a GPU, our two-stage method can ensure the requirement for real-time computation. Codes are released at https://github.com/FLAMEZZ5201/DisGNet.