Introduction  The phenomenon of false water level in the operation of the thermal power unit deep-peak peak operation seriously affects the safe operation of the unit. Accurately predicts the timing and amplitude of the false water level can optimize the control system design.

  Method  Research ideas: the actual water level was regarded as the superposition of the real water level and the false water level. Firstly, the water level model was established according to the energy balance and material balance to calculate the real water level, and then the actual water level was used to reduce the real water level to obtain the false water level,and then used wavelet to analyze the correlation between the fuel quantity, steam drum pressure and other signals and the false water level signal at different time and space scales.

  Result  By analyzing the operation data of the deep peak load section of a 600 MW unit, it is found that the influence of feed water flow and steam flow on the false water level signal is effectively eliminated by mechanism modeling; in the frequency band of 0.062 5 Hz to 0.125 Hz, the fuel quantity and the steam drum pressure are strongly correlated with the false water level signal, respectively.

  Conclusion  It can help determine the occurrence of false water level and its fluctuation range.