Design and implementation of a reconfigurable control architecture for flexible manufacturing teaching units based on EtherCAT
摘要
Traditional flexible manufacturing teaching platforms face significant challenges, including closed control architectures, poor module expandability, and limited compatibility of heterogeneous industrial devices. To address these limitations, this research proposes and implements the EtherCAT Reconfigurable Control Framework for Flexible Manufacturing Teaching Units (EReCT-FM). Integrating reconfiguration requirements across temporal, platform, and application dimensions, EReCT-FM implements a node-oriented device abstraction mechanism and a multi-criteria driver model based on the Weighted Sum Model (WSM) to enable automatic driver loading and binding without manual intervention. To support automated operation, EReCT-FM integrates three core algorithms into a unified control pipeline: EtherCAT auto-configuration for real-time device discovery and network mapping, WSM-based driver selection, and a graph-based dynamic task allocation algorithm for topology-aware system reconfiguration. Using deterministic communication and Distributed Clock synchronization, EReCT-FM enables plug-and-play multi-vendor interoperability and collaborative control of multiple functional modules. Experimental validation demonstrates substantial improvements in reconfiguration efficiency compared with conventional Programmable Logic Controller centric platforms, achieving a single-node detection latency of less than 30 ms; complete multi-node network discovery time scales from 2.9 to 10.9 s depending on the network size, and the synchronization accuracy is maintained in the range of 100 microseconds. This research successfully establishes a scalable and modular cyber-physical architecture for Industry 4.0 smart manufacturing education and interdisciplinary practical training environments.