This study advances the field of intelligent manufacturing by integrating automation, information technology, and digitalization into a novel “five horizontal and three vertical” system architecture. We introduce an enhanced model-free adaptive control (MFAC) algorithm that utilizes a dynamic linearization method alongside pseudo-partial derivatives (PPD) for optimizing data transfer within intelligent manufacturing systems. Using CANoe software for simulation, our research demonstrates the improved MFAC algorithm’s superior bus message transmission efficiency by 0.0224 to 0.026 and a significant reduction in the network load factor by 0.2068. These advancements highlight the efficacy of fieldbus technology in reducing network congestion and enhancing real-time performance in intelligent manufacturing environments.