The invention relates to the technical field of
cavitation monitoring, in particular to a
tidal current energy
water turbine cavitation predicting and monitoring method which comprises the steps that S1, multi-sensor data of a
tidal current energy
water turbine are collected, and the multi-sensor data comprise vibration signals, ultrasonic signals and pressure signals; s2, performing
frequency domain analysis on the vibration
signal, and extracting an energy value of a
high frequency band as a vibration characteristic value; s3, performing transmission attenuation analysis on the ultrasonic
signal, and calculating an
ultrasonic attenuation rate as an ultrasonic characteristic value; s4, performing
frequency domain analysis on the pressure
signal, and extracting a pressure pulsation amplitude at a blade passing frequency as a pressure characteristic value; s5, based on the vibration characteristic value, the ultrasonic characteristic value and the pressure characteristic value, the current
cavitation state is determined through a cavitation state judgment rule; and S6, predicting the evolution trend of the cavitation state, and estimating the residual
safe operation time of the
water turbine under the current working condition. The cavitation recognition result is converted into a quantitative life index, and the method has
engineering application value and safety guarantee capability.