Multi-Level Intelligent Analysis Method for Electromagnetic Transient Simulation Results Based on Supercomputing Platforms

As power systems become increasingly complex, the analysis of large-scale electromagnetic transient simulations faces challenges such as excessive bandwidth consumption for simulation result transmission and insufficient local analysis capabilities of users. To address this bottleneck, this paper proposes a multi-level on-site intelligent analysis framework that runs directly on a supercomputing platform. This framework facilitates the conversion of terabyte-level raw waveform data into the delivery of lightweight core analysis conclusions. The core methodology comprises a three-tier architecture: first, automatically identifying different simulation application scenarios; second, employing embedded zero-tree wavelet and other compression algorithms to achieve efficient data compression while preserving features; and finally, calculating key performance metrics. The efficacy of the framework is demonstrated in this paper through two simulation examples–optimal parameter analysis and impedance sweep analysis–implemented on the MATLAB/Simulink platform.