电厂热工DCS系统数据智能驱动系统设计

为消除电厂热工DCS系统驱动过程人为干扰,提出一种电厂热工DCS系统数据智能驱动系统设计。依据电厂热工系统的运行数据特征,将影响热工系统驱动控制的相关变量,作为驱动模型的初始变量,采用偏最小二乘法处理初始变量之间的相关性和耦合关系,提取特征矩阵;将特征矩阵作为驱动模型的最终输入,并采用最小二乘支持向量机（LSSVM）描述电厂热工系统驱动过程的非线性,实现基于PLS-LSSVM电厂热工DCS系统智能驱动系统设计。实验结果表明,该系统的自重构和回归拟合程度均在0.8以上,并可较好地完成数据降维。

Design of data intelligent drive system for power plant thermal DCS system

In order to eliminate the human interference in the driving process of the thermal DCS system of the power plant,a data intelligent driving system design of the thermal DCS system of the power plant was proposed.According to the operation data characteristics of the thermal system of the power plant,the relevant variables that affect the drive control of the thermal system were taken as the initial variables of the drive model,and the partial least squares method was used to process the correlation and coupling relationship between the initial variables to extract the characteristic matrix;The eigenmatrix was used as the final input of the drive model,and the least squares support vector machine (LSSVM) was used to describe the nonlinearity of the driving process of thermal system in power plant,and the design of intelligent driving system of thermal DCS system in power plant based on PLS LSSVM was realized.The experimental results show that the self-reconstruction and regression fitting degrees of the system were above 0.8,and the data dimension reduction could be completed well.