摘要
针对球磨机出粉量难以测量的问题,文章借助以数据驱动为基础的软测量技术,建立了基于支持向量回归机(support vector regression,SVR)的球磨机出粉量估算模型。为减小模型的误差,使用飞蛾火焰优化(moth-flame optimization,MFO)算法对SVR的惩罚因子C以及径向基函数(radial basis function,RBF)核系数g进行优化。为验证MFO算法的可靠性,将此算法与粒子群优化(particle swarm optimization,PSO)算法、遗传算法(genetic algorithm,GA)进行比较,分别建立了球磨机出粉量的MFO-SVR、PSO-SVR、GA-SVR模型,试验结果表明MFO-SVR估算模型对出粉量有较好的预测和泛化能力。
Aiming at the difficulty of measuring the powder output of the ball mill,with the help of data-driven soft measurement technology,an estimation model of the powder output of the ball mill based on the support vector regression(SVR)machine was established.In order to reduce the error of the model,the moth-flame optimization(MFO)algorithm was used to optimize the penalty factor C of the SVR and the radial basis function(RBF)kernel coefficient g.In order to verify the reliability of the MFO algorithm,this algorithm was compared with the particle swarm optimization(PSO)algorithm and genetic algorithm(GA),and the MFO-SVR,PSO-SVR,GA-SVR models of the ball mill powder output were established respectively,and the test result shows that the MFO-SVR estimation model has a good ability to predict and generalize the powder output.
作者
宋宇
陆金桂
SONG Yu;LU Jingui(School of Mechanical and Power Engineering, Nanjing Tech University, Nanjing 211816, China)
出处
《合肥工业大学学报（自然科学版）》
CAS
北大核心
2021年第10期1347-1352,1362,共7页
Journal of Hefei University of Technology：Natural Science
基金
国家科技支撑计划资助项目(2013BAF02B11)。
