A Metro Cloud Platform Based on Real-Time Processing of Heterogeneous Concurrent Data

With the rapid development of urban transportation, the metro has become a vital means for daily communication. However, during the operation of the metro, a large amount of operational data from various equipment is constantly generated and transmitted to the ground control center, leading to high bandwidth load and poor real-time processing performance of the ground monitoring system. Inspired by the idea of big data processing system, this paper proposes a metro cloud platform for processing large, concurrent, and heterogeneous data generated by metros. We first designed the software structure of the platform according to the overall requirements, including both front-end and back-end systems, and then divided them into several modules with different functions. We employed technologies such as Netty, Kafka, Redis, and TimescaleDB to efficiently handle and store the vast amounts of data generated by metros, and implemented an online playback feature for video stream captured by onboard cameras, which have been processed by deep learning models. Moreover, a data-sending simulator was designed for simulation experiments, and the experimental results verified the functionality and performance of the metro cloud platform.