| With the continuous progress of science and technology,the demand for electricity for national development is increasing day by day,and the construction of energy and power industry is also accelerating day by day.The industrial efficiency and environmental protection has become the core indicator.Low-load thermal power units with backward production capacity are gradually eliminated,and the proportion of high-parameter and high-load thermal power units is increasing continuously.The intelligent development of thermal power plants is constantly promoted by using advanced modern technology,and the operation and production of large thermal power units with low energy consumption,low emission and high intelligence is a research hotspot today.With the increasing degree of automation in thermal power plants,a large number of operation data are generated and stored in the production process of power plants.These data have the characteristics of huge amount,complex structure,various forms,fast increase speed and high value.The ordinary data analysis methods have obvious limitations when dealing with the operation data of power plants.In this paper,big data machine learning technology is used to study and analyze the turbine thermal system of coal-fired power station.According to the actual physical relationship,the appropriate parameters were selected and the BP neural network and the RBF neural network were used to build the heat consumption rate prediction model of the steam turbine unit respectively.After comparison,the RBF neural network prediction model with greater accuracy was finally selected.Then,the quantile impact value(QIV)method was used to weight the inputs of the RBF model,and the prediction model was optimized.Finally,based on the optimized prediction model and particle swarm optimization algorithm,the QIV-RBF-PSO composite optimization model was proposed and built,and the model was tested through an actual case.Using existing advanced technology,large data of electric power production of multi-dimensional,complex analysis,a huge amount of data collection,mining the useful information,make full use of the used for technical reasons and ignore the production information,establish a prediction model of heat consumption rate of big data,without affecting the operation of unit operation parameters in optimization of production,to guide the operation of the field personnel adjustment,The on-line operation optimization of steam turbine unit is realized. |
