摘要
提出了一种新的基于大数据的多入多出系统建模方法。该方法将机理建模、实验建模、智能建模等方法有机结合,通过仿真模型的阶跃实验确定了模型结构和各参数初始范围,挖掘现场运行的历史数据,利用智能优化算法对模型进行校正,得到系统的传递函数模型。解决了不允许或者没有条件在生产现场施加大范围频繁的阶跃扰动实验的问题,确定了模型初始结构,克服了多变量智能寻优时初始值范围不确定的困难。这一新辨识思想成功应用于超超临界机组协调控制系统传递函数辨识,该系统以给水量、给煤量和高调门开度为输入,机组功率、主汽压力和中间点温度为输出,辨识得到了满负荷工况附近的传递函数模型,为协调控制器的设计与优化奠定了基础。
A new MIMO（multi-input multiple-output） system modeling method based on big data was proposed. This method blended the mechanism modeling, experimental modeling, intelligent modeling together, through the simulation model experiments to determine the parameters of the model structure and the initial range of the parameters. Based on the exploiting useful data from the historical operating data of the power plant, using the intelligent optimization algorithm to correction the model, the system’s transfer function was built. It avoided the large-scale frequently step disturbance in the experiments production site. The initial structure of the model was determined, and the problem of multi-variable optimization intelligent’s initial value range was solved. This new identification idea was successfully used to the identification of the ultra-supercritical unit coordinated control system. In this model, total water flow, fuel command and turbine governor valve position were considered as inputs and turbine power, main steam pressure and intermediate point temperature were considered as outputs. Finally, the identification results have been given near full load conditions which can be reference of the design or optimization of the coordinated control system.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2014年第7期1454-1459,1510,共7页
Journal of System Simulation
关键词
建模方法
多变量系统
数据挖掘
协调控制系统
modeling method
multivariable system
data mining
coordinated control system
