Dynamic NOx Emission Prediction Based on Sandcat Swarm Optimization-least Squares Support Vector Machine
Aiming at the problem that the frequent peak shaving of thermal power units leads to the unstable combustion state of the unit,causing a large fluctuation range of NOx generation at the boiler outlet,a dynamic NOx emission model based on Sand Cat Swarm Optimization(SC SO)and Least Squares Support Vector Machine is proposed.First,the k-nearest neighbor mutual information is used to calculate the time delay and filter auxiliary variables.Then,based on SCSO algorithm,the order of input variable is selected.Using the information including the time delay and order of auxiliary variables as the input of the model,the SCSO algorithm optimizes the parameters of the least squares support vector machine and establishes the least squares support vector machine prediction model of dynamic NOx emission(SCSO-LSSVM dynamic soft sensing model).Finally,the model is compared with the LSSVM model without delay,the LSSVM model with delay and the dynamic soft sensor model with Particle Swarm Optimization algorithm to optimize the parameters of least squares support vector machine.The results show that,compared with other models,the SCSO-LSSVM dynamic soft sensor model established in this paper has the smallest root mean square error,the smallest mean absolute error and the smallest mean absolute error,while it has the highest prediction accuracy.Moreover,it can also predict the NOxproduction well when the NOx production fluctuates sharply,and has good dynamic characteristics.
NOx concentrationk nearest neighbor mutual informationsandcat swarm optimization algorithmleast squares support vector machinesoft sensor model
国家科技期刊平台
NSTL
万方数据
