Abstract：Due to the limitation of hardware platform, the current fault recorder terminal has limited ability to deal with a large amount of transient and steady data in the secondary circuit of converter station control and protection, and the possibility to improve the analysis accuracy rate only by software algorithm is limited. At present, the identification accuracy rate can not meet the requirement of reliable application. When there is a fault, the operation and maintenance personnel take the corresponding waveform to check, but for a large number of monitoring waveforms, it is impossible to identify the small changes of electrical parameters manually, which reduces the work efficiency of the operation and inspection personnel. In this paper, a real-time monitoring data analysis and early warning system based on AI edge calculation is proposed. The AI acceleration module is configured at the fault recorder device end, and the multi-source data model of converter station control and protection is built. The multi-dimensional data processing algorithm of statistical analysis and waveform analysis is used to calculate the real-time data of secondary circuit of control and protection of converter station, so as to discover the concealed defects of control and protection secondary circuit in time and warn early.
周源, 鞠翔, 刘英男, 李亚锦. 基于AI加速模块的换流站控制保护二次回路预警系统[J]. 电气技术, 2021, 22(5): 62-67.
ZHOU Yuan, JU Xiang, LIU Yingnan, LI Yajin. Early warning system of control and protection secondary circuit of converter station based on AI acceleration module. Electrical Engineering, 2021, 22(5): 62-67.
2021, Vol. 22 
