摘要
为了提高生产效率与能源利用率,提出了采用最小二乘支持向量机(Least squares support?vector machines LSSVM)算法对啤酒厂煮沸车间蒸汽量的消耗预测.根据历史数据样本,采用RBF核函数作为LSSVM的核函数,交叉验证结合网格搜索来优化参数,用所建立决策函数作为预测模型.实验结果表明,数据样本的预测集和测试集的均方差均达到0.006,拟合相关参数达到99%以上,LSSVM方法能够快速准确的预测该车间在生产旺季的蒸汽消耗,为企业节能控制方案的制定提供了理论依据.
Least squares support vector machines（ LSSVM） algorithm was proposed to predict the steam consumption in the boiling workshop of the brewery to improve production efficiency and energy utilization ratio. According to samples of historical data,the RBF kernel function was used as the kernel function of LSSVM and cross validation combined with grid- search was used to optimize the parameters,with the established decision function being the forecasting model. Experimental results show that MSE of the training set and prediction set reaches 0. 0060; correlation coefficients reached above 99%. LSSVM method can quickly and accurately predict the energy consumption of the plant in the busy season,providing a theoretical basis for the enterprise energy saving control scheme.
出处
《内蒙古科技大学学报》
CAS
2015年第4期341-343,共3页
Journal of Inner Mongolia University of Science and Technology
基金
国家科技型中小型企业技术创新资助项目(12C26212201322)
关键词
最小支持向量机
能耗预测
交叉验证
网格搜索
Least squares support vector machines（LSSVM）
Energy consumption forecasting
cross validation
grid search
