文章基本信息

标题：RESEARCH ON APPLICATION OF LIMIT LEARNING METHOD AND LEAST SQUARE SUPPORT VECTOR MACHINE IN ECOLOGICAL EMISSION PREDICTION
本地全文：下载
作者：Guobin Chen ; Jun Liu ; Gang He 等
期刊名称：Fresenius Environmental Bulletin
印刷版ISSN：1018-4619
出版年度：2019
卷号：28
期号：12
页码：9223-9229
语种：English
出版社：PSP Publishing
摘要：The discharge of ecological environment is an important index to restrict economic development. Effective controlling and optimizing the discharge of ecological environment is an important measure to protect the ecological environment at present. In order to accurately measure NOx emission from pulverized coal fired boilers, a maximum learning method (ELM) was proposed to optimize least squares support vector machine (LSSVM) for predicting ecological emission. To overcome the shortcomings of over-fitting of traditional extreme learning machine, it was proposed a novel NOX emission prediction model (ELM-LSSVM) based on the combination of extreme learning machine and least squares support vector machine (LSSVM). This method used the improved maximum learning method to optimize the LSSVM parameters, established the optimization model, and predicted the different NOx emissions of coal-fired boilers. The learning samples of NOx traffic were obtained by phase space reconstruction, and the least squares support vector machine was introduced to improve the limit learning. The training set of NOx traffic was learned, and the performance of the model was tested by simulation experiments. Finally, taking 330MW coal-fired boiler of a thermal power plant as the research object, the NOx prediction model of ELM- LSSVM was established. The simulation results showed that the soft sensor model has higher prediction accuracy and generalization ability, and it can effectively predict NOx emission.
关键词：Ecological environment;ELM;LSSVM;optimization model