Research on Online Evaluation Method of Error State of All-Fiber Current Transformer Based on Improved MWPCA

All-fiber current transformer is widely used in high-voltage power system because of its good insulation performance, strong electromagnetic interference resistance and no magnetic saturation. However, the error stability of all-fiber current transformer is poor in long-term operation, and the traditional error period detection method based on the standard has the risk of lagging behind the actual situation. In order to solve this problem, this paper proposes an improved MWPCA on-line error state evaluation method for all-fiber current transformers: the measurement data of three redundant all-fiber current transformers at the same measuring point in converter station are collected, and the error evaluation is carried out by using the improved MWPCA method proposed in this paper. The experimental results show that the model proposed in this paper has good detection effect under FOCT fixed error and gradual error, and meets the requirements of 0.2-level detection accuracy. At the same time, compared with the traditional PCA method, the method proposed in this paper also reduces the misjudgment rate when the error fluctuates normally.