Simulation Study On Off-Design Performance Of Combined Cycle Power Plant Based On Data-Driven

In iron and steel enterprises,the efficient utilization of blast furnace gas is an effective way to achieve energy saving and carbon reduction.Because of its high efficiency,environmental protection and operation flexibility,the low calorific value gas-steam combined cycle unit has become the best choice for efficient recycling of blast furnace gas.However,in the actual operation process,environmental changes and gas supply fluctuations often cause the unit to fail to operate efficiently under design conditions.Therefore,it is of great significance to establish a simulation model of the unit under variable operating conditions,calculate the key operating parameters and cycle efficiency of the unit under actual operating conditions,and explore the optimal operation strategy of the unit.In this thesis,the simulation modeling of a 300 MW gas-steam combined cycle generator set put into production is studied.Firstly,the mechanism model of the combined cycle unit is studied,and the variable condition simulation model of key equipment such as gas compressor,air compressor,gas turbine,waste heat boiler and steam turbine is constructed to explore the relationship between the relevant physical quantities of the main equipment under variable operating conditions.However,due to the complex internal structure and nonlinear quantitative relationship of the combined cycle generator set,the simulation model based on the mechanism cannot accurately predict the operation status of the unit.Therefore,this thesis uses the BP neural network with strong nonlinear mapping ability to model the combined cycle generator set.Based on the analysis of the characteristics of the combined cycle generator set,the structure of the BP neural network model of the gas-steam combined cycle generator set is established through continuous experiments using a large number of historical operation data,and the parameter configuration is optimized.After testing,the trained network can well reflect the operating characteristics of the unit under variable operating conditions,and the prediction error can be stably controlled within 2 %,which lays a good foundation for subsequent research.Based on the BP neural network model,the influence of ambient temperature change and gas calorific value fluctuation on the operation performance and operation scheme of the combined cycle unit under the hybrid control strategy of fuel regulation and IGV is discussed.The specific contents are as follows :(1)Under the hybrid control strategy,the influence of ambient temperature change on the performance of combined cycle is analyzed.The calculation results show that different control methods will make the combined cycle show different rules under the influence of ambient temperature.When the gas turbine power is 0.5 and 0.9,the unit adopts the control method of single fuel quantity,and the combined cycle generator set will obtain higher operating efficiency when the ambient temperature is 0 °C.When the relative power of the gas turbine is 0.7,the unit adopts a hybrid control mode of IGV and fuel regulation,and the combined cycle generator set will obtain higher operating efficiency at an ambient temperature of 30 °C.(2)When analyzing the change of calorific value,three load operating conditions of gas turbine with relative power of 0.5,0.7 and 0.9 are selected for analysis.The calculation results show that the increase of fuel calorific value will reduce the amount of fuel required for unit operation,but the operation efficiency of gas turbine and combined cycle unit will decrease.With the increase of unit load,the reduction range will increase,and the calorific value of gas will increase by 500 k J / Nm3.When the relative power of gas turbine is 0.9,the efficiency of combined cycle will lose 2.9 %more than that of 0.5.