Prediction On Maximum Output Of CFB Thermal Power Plants Burning Variable Coals

The complicated grid structure and system of Mengxi Power Grid make it very difficult to dispatch and control the power grid.Mengxi Power Grid has the largest CFB machine assembly capacity in the country,but the frequent changes in coal types and the restrictions on auxiliary machines will limit the maximum output of the unit.At present,there is little research on the prediction of the maximum output of units with variable coal quality,and the research on the prediction of the maximum output of CFB units under the condition of changing coal quality is of great significance to ensure the secure and steady operation of the power grid.Taking a CFB unit as the research object,first of all,a large amount of data stored in the power plant system is used to remove abnormal data using the Laida criterion,and then the data is analyzed and calculated using the average impact value method that can directly reflect the degree of influence of the respective variables on the dependent variable.After processing and analyzing the data provided by the power plant system,including 56 independent variables and1 dependent variable,it is found that the main steam flow is the decisive factor that affects the output of the unit,and the influence weight reaches 0.411.For different coal types,the overall thermal calculation of the existing boilers based on the verification method is performed to predict the main steam flow of the existing boiler based on the verification method is performed to predict the main steam flow of the existing boiler burning different coal types.Use Matlab for programming,divide the entire calculation process into structural calculation module,fuel calculation module,flue gas characteristic module generated by combustion,flue gas enthalpy temperature meter module,water vapor enthalpy value and temperature calculation module,boiler heat balance module,heat exchange calculation module and alarm program module.According to the flue gas and steam water flow of the boiler,the program calculation and related modules are called,and finally the exhaust gas temperature is obtained to obtain the thermal efficiency of the boiler,and the main steam flow of the boiler is predicted.The program is checked with the coal currently used in the boiler,and the flue gas temperature and soda water temperature at the entrance and exit of each node are compared,and the error is found to be within ±2%,confirming the reliability of the program.The BP neural network is used to model the unit’s output prediction,the network structure is determined,and the genetic algorithm is used to optimize the BP neural network to reduce errors.The experiment compares and analyzes the relative error and root mean square error of the output value and the expected value under different hidden layer neurons.The data shows that the comprehensive results are better when 7 hidden layer neurons are selected.Taking the relevant data stored in the unit as an example,the results show that the relative error between the output value obtained from the model test and the expected value is about ±2%.The main steam flow obtained by the boiler thermal calculation program is input into the neural network model,which can realize the maximum output prediction of the unit’s changing coal quality.