Research On Comprehensive Evaluation Method Of Operation State Of MMC-MTDC System

Multi-Terminal Flexible DC Transmission System(MTDC)based on Modular Multilevel Converter(MMC)has good scalability,flexible control,high transmission efficiency,and harmonic content Low-grade advantages,in recent years began to be applied in the field of transmission engineering at home and abroad.However,the research in this field at this stage mainly focuses on the mechanism analysis and corresponding control strategies of sub-modules and AC/DC line faults in the station.The related research on the evaluation of MMC-MTDC operation status is still in its infancy,lacking the MMC-MTDC transmission The system effectively extracts,summarizes and analyzes the corresponding operating state characteristics under different working conditions and control strategies,and further realizes the inversion and identification of AC transient faults according to its operating state,so an accurate state evaluation method is urgently needed:Study the change of MMC-MTDC operating state to provide a basis for judgment.This article first summarizes the topology,working principle and operating characteristics of MMC,and establishes the dynamic mathematical model of the AC and DC sides of the MMC converter station;analyzes different operating conditions,control strategies,and AC side transient faults on the MMC-MTDC system The impact;built a five-terminal MMC-MTDC transmission system model based on OPT-5600 to provide theoretical support for subsequent research.Then,by simulating the bridge arm voltage signals of MMC converter stations in different operating conditions,and intercepting the effective period in the time domain as the sample signal to be studied;based on the transient time-varying signal relative to the time-invariant signal in the time-frequency domain Out of the particularity,the optimal group decomposition method is proposed.The optimal group decomposition energy method is used to decompose the electrical information in the MMC converter station.On the basis of the decomposition of the dominant oscillation components,the energy entropy is calculated to establish all samples.Feature vector.Finally,using the results of feature extraction to build an indicator set,since the improved analytic hierarchy process is used to judge the weight of each index subject to the subjectivity of experts when evaluating the operating status of the MMC-MTDC system,the improved analytic hierarchy process and anti-non-extended entropy weight method are combined An improved analytic hierarchy process-anti-non-extended entropy evaluation method was constructed,and the operation status evaluation of the MMC-MTDC system was completed.Based on the results of state evaluation and grading,the support vector machine is used to realize the inversion and identification of AC transient fault types.The radial basis kernel function is selected by LIBSVM and the parameters are optimized by cross-validation method.A multi-class support vector machine model is established to identify AC transient fault,the simulation results show that this method is effective.