Study on prediction of operational data of turbine in TRT system based on artificial neural network

Blast furnace top gas recovery turbine unit, TRT for short, is an energy saving system. Turbine is the key equipment in TRT system. The normal operation of turbine is directly related to the capacity of TRT system and the benefit of enterprises. In order to achieve the prediction of operation data of turbine, this paper firstly processed the turbine operation data collected from production scene with the soft and hard threshold compromise algorithm in decomposing the wavelet coefficient, and filtered the noise. Then, this paper established the prediction model of BP neural network, adjusted the structure of the neural network and trained it for the prediction results. The result shows that on the premise of filtering the noise with wavelet transform, the BP neural network can achieve the prediction of turbine operation data in TRT system effectively.