Thermal power unit air compressor energy saving research based on compressed air flow prediction
To address the problems of scattered air compressor layout and frequent loading and unloading in thermal power units,a method is proposed to optimize the combination of air compressor network operation based on compressed air flow prediction results.In order to improve the prediction accuracy of the prediction model,a prediction model(POA-LSTM)combining Pelican Optimization Algorithm(POA)and Long Short-Term Memory Neural Network is proposed.Through correlation analysis,the air compressor operating current,storage tank pressure,air compressor exhaust pressure and air compressor exhaust temperature are selected as the influencing factors of compressed air flow,and the prediction results show that the prediction error rate of the model is less than 4%,which can more The prediction results show that the prediction error rate of the model is less than 4%,which can predict the change trend of compressed air flow more accurately.Based on the prediction results of air consumption,we carried out the renovation test of miscellaneous compressed air network and optimized the combination operation of air compressors,which can reduce one miscellaneous air compressor and reduce energy consumption by 13.01%.The networked air supply method based on compressed air flow prediction can save more than 4.4 million kilowatt hours of electricity,2.68 million yuan of electricity cost and 2559 tons of carbon dioxide emission for the whole plant.
compressed air flow predictionLSTMpelican optimization algorithmair compressor energy saving
万方数据
维普
