基于改进最优觅食算法的锅炉NOx排放预测研究

Prediction of NO x Emission of Boiler Based on Improved Optimal Foraging Algorithm

提出了一种改进的最优觅食算法(POFA),在最优觅食算法中引入自适应惯性权值与全局最优解来改进算法的更新公式,同时加入相空间搜索的机制。利用改进的最优觅食算法优化极端学习机(ELM)构建一个改进的极端学习机模型(POFA-ELM),并用该模型对锅炉NOx的排放特性进行建模。将该模型与ELM、差分进化算法、粒子群算法、人工蜂群算法以及基本的最优觅食算法优化的ELM模型进行比较。结果表明:该模型的预测精度更好,泛化能力更强,可以更加准确地预测NOx的排放质量浓度。

An improved optimal foraging algorithm is proposed,the update direction of adaptive inertia weight and global optimal solution is introduced in the optimal foraging algorithm,at the same time,the mechanism of phase space search is introduced.The improved optimal foraging algorithm(POFA)optimized by extreme learning machine(ELM)constructs an improved extreme learning machine model(POFA-ELM)and uses the POFA-ELM model to model the boiler NO x emission characteristics.Compared with ELM and differential evolution algorithm,particle swarm optimization,artificial bee colony algorithm,and basic optimal foraging algorithm optimized ELM model,the results show that the POFA-ELM model has better prediction accuracy and stronger generalization ability,which can predict the NO x emission quality concentration more accurately.