Real-time metering data processing platform

The invention relates to the technical field of real-time analysis, in particular to a real-time measurement data processing platform which comprises a data modeling module, a probability inference module, an anomaly detection module, a topology analysis module, a correlation analysis module, a chaos prediction module, a graph embedding analysis module and a swarm intelligence optimization module. According to the method, through the dynamic Bayesian network model and the Bayesian reasoning algorithm, the accuracy of the platform in modeling time sequence data and risk assessment is remarkably improved, the irregular pattern is efficiently identified by using a statistical model in anomaly detection, assessment of potential risk points is enhanced, and the accuracy of risk assessment is improved. Topology analysis and a hypergraph theory deepen key feature extraction and multivariate relation analysis, data structure insight is enhanced, a chaos theory and a graph embedding technology are combined, complex dy