Abstract: With the rapid development of industrial IoT, monitoring systems face challenges integrating numerous devices using different protocols. This paper proposes a monitoring system architecture supporting automatic mounting of acquisition equipment, using hashtables to construct data structures and shared memory for fast data exchange and low-latency forwarding. Experiments show CPU utilization reduced by 22% and memory usage by 43%, with applications in multiple new energy monitoring systems.