IGSA-LSSVM软测量模型预测燃煤锅炉NOx排放量

摘要

提出一种基于改进引力优化算法(IGSA)优化最小二乘支持向量机(LSSVM)的软测量模型(IGSA-LSSVM)以精确测量煤粉锅炉NOx排放量.首先,针对引力搜索算法易陷入局部最小、全局优化能力差的问题,提出了一种改进的引力搜索算法,利用网格算法初始化种群,基于适应度值自适应递减惯性权重更新质点位置以提高全局优化性能;然后,采用IGSA优化选择LSSVM的超参数以改善模型的预测精度和泛化能力;最后,以330 MW燃煤锅炉为研究对象,建立IGSA-LSSVM的NOx软测量模型,仿真结果表明该软测量模型具有更高的预测精度和泛化能力,能有效测量NOx排放量.%Based on least squares support vector machine optimized by improved gravitational search algorithm (IGSA-LSSVM),an intelligent soft sensing method to accurately measure the NOx emission of the coal-fired boiler is presented.Firstly,the GSA have the drawbacks of easy to fall into local minimum and poor global search ability,so an improved version of GSA is proposed to improve global optimal performance,using the grid algorithm that is employed to initialize the population and using the adaptive decreasing inertia weight based on fitness value of optimization problems that is introduced into position update.Secondly,IGSA is developed to find the optimal parameters of LSSVM to improve the regression accuracy and generalization ability for predicting NOx emission.Finally,a soft computing method based on IGSA-LSSVM is established to forecast NOx emission of a 330 MW coal-fired boiler.The simulation results show that the IGSA-LSSVM model demonstrates better regression precision and generalization capability,it can accurately measure NOx emission.