文献类型：会议论文

英文题名：VAST: Architecture for Low-Cost Rapid Control Prototyping Systems

作者：Luo, Qi[1];Qian, Zhiqin[1];Wang, Chen[1];Zeng, Liding[1];Jiang, Xuejiao[1];Xu, Lvqi[1];Liu, Wei[1];Peng, Chen[1]

机构：[1]East China Univ Sci & Technol, Dept Mech & Power Engn, Shanghai, Peoples R China

会议论文集：11th International Conference on Electrical Engineering, Control and Robotics-EECR

会议日期：APR 18-20, 2025

会议地点：Changzhou, PEOPLES R CHINA

语种：英文

外文关键词：rapid control prototyping; large language model; control experiment; low-cost; code generation

摘要：This paper proposes an architecture named VAST for constructing low-cost rapid control prototyping (RCP) systems. With the cross-language code generation capability of the large language model (LLM), VAST significantly reduces the hardware cost and software development complexity of the RCP systems while maintaining adaptability to general control algorithm testing tasks. Under this architecture, researchers conducting control algorithm validation experiments can build customized RCP systems for their specific testing task in a shorter time and at a lower cost. In this paper, a specific RCP system under VAST architecture is constructed using an STM32 microcontroller along with the LLM and the open-source signal visualization software VOFA+. The proposed architecture is validated through a trajectory tracking control experiment on a controller based on a radial basis function neural network (RBFNN) which is applied to a two-degree-of-freedom (DOF) planar five-bar parallel robot. Experimental results confirm the effectiveness of the VAST architecture in facilitating efficient and cost-effective control experiment validation.