Pumped-storage unit abnormality identification and fault alarming model based on Chebyshev inequation

In order to solve problem on setting alarming values in pumped-storage unit status monitoring, working load point distributions of 14 power stations have been analysed. The statistics indicate that all the units run in constant power mode under most conditions, and the peak-to-peak values of run-outs and vibrations comply with normal distributions after the hypothesis test. Meanwhile the distribution of data has weak relation with water head. Based on the results obtained, parameter-adjustable abnormality identification model has been established by Chebyshev inequation, and pumped-storage unit status monitoring and alarming strategy for different working point has been fulfilled.