基于PSO-Elman模型的火电厂烟气含氧量预测  

Prediction of the Oxygen Content in Flue Gas of Power Plant Based on PSO-Elman Model

作　　者：张玮 陈超波[1] 王景成 李继超[1] 郝爽洁[1] ZHANG Wei;CHEN Chao-bo;WANG Jing-cheng;LI Ji-chao;HAO Shuang-jie(College of Electronic Information Engineering,Xi’an Technological University,Xi’an 710021,China;Department of Automation,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]西安工业大学电子信息工程学院,西安710021 [2]上海交通大学自动化系,上海200240

出　　处：《自动化与仪表》2020年第4期75-79,85,共6页Automation & Instrumentation

基　　金：陕西省工业领域重点研发计划项目(2018ZDXM-GY-168)。

摘　　要：针对火电厂烟气含氧量测量精度较低、测量成本较大等问题,提出基于PSO-Elman网络模型的烟气含氧量预测方法。选择合理的相关辅助变量,引入Elman神经网络建立辅助变量与烟气含氧量的关系模型;利用PSO对Elman中所有的权值、阈值进行寻优,将其最优权值和阈值作为初值赋给Elman;经过训练,建立基于PSO-Elman模型的预测模型,完成火电厂烟气含氧量的软测量。通过仿真,与Elman网络和LSSVM模型的预测结果作对比,所提模型具有较高的预测精度和良好的泛化能力。Aiming at such problems of inaccuracy and large costs in current measuring the oxygen content in flue gas of power plant,a prediction model combining Elman neural network and Parti cle Swarm Optimization(PSO-Elman)was proposed. Select reasonable correlative variables and the Elman neural network is introduced to establish the prediction model of relevant variables and oxygen content in flue gas. Use PSO algorithm to optimize the weights and thresholds of the Elman neural network which would be assigned to this neural network. By training,the prediction model of the oxygen content in flue gas of power plant is established based on PSO-Elman network model. Through simulations of specific examples,this prediction model is proved that it has a high accuracy as well as generalization ability in the operation of the power plant being compared with the prediction result of the Elman network and the LSSVM model.

关 键 词：烟气含氧量 火电厂 PSO算法 ELMAN神经网络 预测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TM621.7[自动化与计算机技术—控制科学与工程]