论文部分内容阅读
为克服现有机理建模和试验建模方法的不足,从基本建模理论出发,研究了利用现场数据对热工对象进行建模的问题。基于现场数据建模需要对热工过程输入信号的可激励性、控制系统闭环可辨识性、多变量系统输入信号的无关性等进行判断。结果表明:在较大的机组负荷波动过程中采集的运行数据能够满足输入信号的可激励性;热工过程的大滞后现象对保证热工控制系统的闭环可辨识性有利;含有噪声的实测信号能够保证多变量系统的可辨识性。在满足以上条件的情况下,可以基于现场数据对热工对象进行建模研究。
In order to overcome the deficiencies of existing mechanism modeling and experimental modeling methods, starting from the basic modeling theory, the problem of modeling thermal objects by using field data was studied. Based on the field data modeling, it is necessary to judge the excitability of the input signal of the thermal process, the recognizability of the closed loop of the control system and the independence of the input signal of the multivariable system. The results show that the operating data collected during larger unit load fluctuation can satisfy the stimulability of input signal. The large hysteresis of thermal process is beneficial to ensure the closed loop recognizability of thermal control system. The measured signal with noise Multivariate system can ensure the identifiability. In the case of meeting the above conditions, the thermal engineering can be modeled based on the field data.