A multi-source heterogeneous massive operation and maintenance data collection method for substations based on cloud-edge collaboration and artificial intelligence

Real-time collection of operation and maintenance data enables timely identification and resolution of potential faults, thereby ensuring the safe and stable operation of power grids. However, achieving both real-time and secure data acquisition remains a persistent challenge in this field. This paper proposes a cloud-edge collaboration and artificial intelligence-based method for collecting multi-source heterogeneous massive operation and maintenance data in substations. At the terminal layer, substation equipment terminals gather multi-source heterogeneous massive operation and maintenance data and transmit them as messages via diverse communication protocols to the nearest edge server. At the edge layer, the received data are aggregated according to the optimal edge server determined by the cloud layer, thereby completing the collection of massive operation and maintenance data. In the cloud layer, the aggregation task offloading unit employs an artificial intelligence-based Particle Swarm Optimization (PSO) algorithm. With the objective of minimizing total delay and total energy consumption, the algorithm derives data aggregation task offloading decisions to ensure the efficient execution of operation and maintenance data collection tasks. Experimental results show that the proposed method exhibits relatively small fluctuation amplitudes in both uplink and downlink communication links, indicating stable data transmission speed and quality. The maximum latency is approximately 3.4 s and 2.4 s, respectively, while data integrity remains between 94 and 98%, and resource utilization stays between 5 and 9%, confirming its practical effectiveness.