基于PCA和支持向量机的电站入炉煤量软测量技术
|
摘要
电厂燃煤机组中,锅炉的入炉煤量直接关系到炉膛的燃烧情况,提供锅炉需要的煤粉量对提高锅炉的燃烧效率有重要意义。然而受到各种变化因素的影响,造成入炉煤量的测量误差较大。针对入炉煤量的测量问题,本文提出将主元分析(PCA)技术与支持向量机(SVM)相结合,建立入炉煤量的软测量模型,该方法利用主元分析技术将建模数据进行压缩,降低了支持向量机建模的难度,提高入炉煤量计算的可靠性和准确性。电厂实际运行数据验证表明:该方法能有效跟踪入炉煤量的变化,且计算简便,具有较好的推广应用价值。
In the power plant coal-fired units, the as-fired coal quantity of the boiler directly related to the combustion of the boiler. The appropriate amount of pulverized coal is of great significance for improving the combustion efficiency of the boiler. However, a large measurement error of the as-fired coal quantity caused by the influence of various factors is exited. For the measurement of as- fired coal quantity, this paper proposes the soft sensor method. This method integrates principal component analysis(PCA) and support vector machine(SVM), then establishes the model of the as-fired coal quantity, it uses principal component analysis to the modeling data compression, reduces the modeling difficulty of support vector machine, improves the reliability and accuracy of as- fired coal quantity. The actual operation data in the plant verifies the soft sensor method can effectively track the change trend of as-fired coal quantity, its calculation is simple, and has a better promotional and application value.
In the power plant coal-fired units, the as-fired coal quantity of the boiler directly related to the combustion of the boiler. The appropriate amount of pulverized coal is of great significance for improving the combustion efficiency of the boiler. However, a large measurement error of the as-fired coal quantity caused by the influence of various factors is exited. For the measurement of as- fired coal quantity, this paper proposes the soft sensor method. This method integrates principal component analysis(PCA) and support vector machine(SVM), then establishes the model of the as-fired coal quantity, it uses principal component analysis to the modeling data compression, reduces the modeling difficulty of support vector machine, improves the reliability and accuracy of as- fired coal quantity. The actual operation data in the plant verifies the soft sensor method can effectively track the change trend of as-fired coal quantity, its calculation is simple, and has a better promotional and application value.
引文
[1]孙家鼎,任合社.在线实时监测锅炉给粉量的新方法[J].中国电力,2009,42(7):72-74.
[2]吴永存,陈卫.基于机器学习的火电机组一次风量软测量技术[J].电站系统工程,2014,30(4):61-64.
[3]阎威武,朱宏栋,邵惠鹤.基于最小二乘支持向量机的软测量建模[J].系统仿真学报,2003,15(10):1494-1496.
[4]郑小霞,钱锋.基于PCA和最小二乘支持向量机的软测量建模[J].系统仿真学报,2006,18(3):739-741.
[5]王华忠,俞金寿.基于PCA-SVM的软测量建模方法与应用[J].自动化仪表,2004,25(2):16-19.
[6]俞金寿.软测量技术及其应用[J].自动化仪表,2008,29(1):1-6.
[7]Brosillow.Inferential control of process[J].AICh EJ,1978,24(3):485-509.
[8]朱学峰.软测量技术及其应用[J].华南理工大学学报,2002,30(11):61-66.
[9]邵信光,杨慧中,陈刚.基于粒子群优化算法的支持向量机参数选择及其应用[J].控制理论与应用,2006,23(5):740-748.
[10]王建峰,张磊,陈国兴等.基于改进的网格搜索法的SVM参数优化[J].应用科技,2012,39(3):28-31.
[2]吴永存,陈卫.基于机器学习的火电机组一次风量软测量技术[J].电站系统工程,2014,30(4):61-64.
[3]阎威武,朱宏栋,邵惠鹤.基于最小二乘支持向量机的软测量建模[J].系统仿真学报,2003,15(10):1494-1496.
[4]郑小霞,钱锋.基于PCA和最小二乘支持向量机的软测量建模[J].系统仿真学报,2006,18(3):739-741.
[5]王华忠,俞金寿.基于PCA-SVM的软测量建模方法与应用[J].自动化仪表,2004,25(2):16-19.
[6]俞金寿.软测量技术及其应用[J].自动化仪表,2008,29(1):1-6.
[7]Brosillow.Inferential control of process[J].AICh EJ,1978,24(3):485-509.
[8]朱学峰.软测量技术及其应用[J].华南理工大学学报,2002,30(11):61-66.
[9]邵信光,杨慧中,陈刚.基于粒子群优化算法的支持向量机参数选择及其应用[J].控制理论与应用,2006,23(5):740-748.
[10]王建峰,张磊,陈国兴等.基于改进的网格搜索法的SVM参数优化[J].应用科技,2012,39(3):28-31.