Fault Diagnosis and Data Reconstruction of Transmission Lines Monitoring Sensors

Recently, the development of power systems with high penetration renewable energy, the requirements for transmission lines omnidirectional perception are becoming higher and higher. Under the background of increasingly complex operation, weather etc., the performance of on-line monitoring sensors for transmission lines in practical application is prone to degradation. And the growth in service time and other factors also lead to the sensor’s deterioration or even failure. Therefore, this paper proposed a sensor failure diagnosis, classification and data reconstruction method. Firstly, the repetitive uncertainty is defined and applied to evaluate the performance of sensors. Then, the extreme learning machine is applied to diagnose and classify the sensor faults. Finally, a data reconstruction method combining with the proposed repetitive uncertainty is proposed for monitoring sensors of transmission lines. The simulations are performed and the results show that the proposed method can diagnose the failure sensor and reconstruct the measurement data effectively.