PREDICTION OF INDIRECT AIR COOLING BACK PRESSURE AND CONTROL OPTIMIZATION OF VARIABLE FREQUENCY CIRCULATING PUMP
According to the actual operation data related to back pressure of 2 × 350 MW circulating fluidized bed unit 1 of Hepo power plant in Shanxi Province from July 15 to August 15,2019,by analyzing the influencing factors(ambient temperature,unit load,condenser temperature,pressure,etc.),the correlation coefficient method and principal component analysis method were used to reduce the dimension of data.On the Python platform,Keras,TensorFlow and other libraries were used to compile the prediction algorithm,and the back pressure prediction model was established.The RNN neural network was used to predict the back pressure,and we analyzed the prediction results.The model of variable frequency circulating pump was identified combining with the actual operation data,and the fuzzy PID was used to optimize its control.
Indirect air coolingBack pressure predictionVariable frequency circulating pumpLSTMPython
国家科技期刊平台
NSTL
万方数据
维普
