论文部分内容阅读
支持向量机(SVM)方法以统计学习理论作为其理论基础,采用结构风险最小化原则,具有精度高、全局最优及泛化性好等特点。建立了一个基于支持向量机的火电机组性能在线计算模型,该模型将机组供电标准煤耗率与其影响因素之间复杂的非线性关系通过训练样本构建函数很好地表现出来,可对机组运行经济性进行准确在线计算。将该模型应用于300MW火电机组中,供电煤耗率仿真的最大相对误差均小于0.2%,结果表明SVM模型能够精确地计算火电机组的经济性,是一种有效的模型。
Support Vector Machine (SVM) method takes statistical learning theory as its theoretical foundation, adopts the principle of minimizing structural risk, and has the characteristics of high precision, global optimum and generalization. An on-line calculation model of thermal power plant performance based on SVM is established. This model shows the complex nonlinear relationship between the standard coal consumption rate of power unit and its influencing factors through the training sample construction function, Accurate online calculation. Applying the model to the 300MW thermal power unit, the maximum relative error of the coal consumption rate simulation is less than 0.2%. The results show that the SVM model can calculate the economy of the thermal power unit accurately and is an effective model.