基于PSO参数优化的SVM飞灰含碳量建模研究

Research on modeling of SVM carbon content in fly ash based on PSO parameters optimization

为了解决火力发电厂飞灰含碳量实时监测和控制难题,笔者根据支持向量机(SVM)模型参数,建立了基于粒子群算法(PSO)参数优化的支持向量机模型,并用其对某电厂的锅炉飞灰含碳量进行实时监测。监测结果表明,基于粒子群算法参数优化的支持向量机监测模型较常规模型有着良好的性能,在线监测精度高,可使电厂有效监测和控制飞灰含碳量。

In order to solve the difficulties in real- time monitoring and control of carbon content in fly ash in thermal power plant,the author established the model of support vector machine based on PSO parameters optimization according to SVM model parameters,and online monitored the carbon content in fly ash. The monitoring result shows that the monitoring model of support vector machine based on PSO parameters optimization enjoys the better performance than the common models do with higher accuracy of online monitoring which enables effective monitoring and control of carbon content in fly ash.