摘要
管式加热炉是石化企业的耗能大户,节能优化操作一直是研究的热点。合理控制烟气出口氧含量与温度,能够有效地降低加热炉的能耗,就需要建立精确的加热炉数学模型。但由于管式加热炉是一类复杂的多变量系统,具有滞后大,干扰因素多,耦合性强,系统存在闭环等特点,若采用一般的子空间方法会存在有偏估计的问题。现采用一种改进的子空间辨识算法,通过设计辅助变量,能够有效解决原有方法辨识结果有偏的问题。通过将上述方法应用到一个仿真实例和某企业的管式加热炉模型的辨识中,并与原有子空间方法辨识结果进行了比较。仿真结果证明,改进方法拥有更高的辨识精度,更加适合于复杂的工业环境,为加热炉优化设计提供了参考。
Tubular furnace is the major energy consumption in petrochemical enterprises,the energy-efficient and optimization operation has been the research focus. The energy consumption of the furnace can be reduced effectively by the reasonable control of outlet flue gas oxygen content and the temperature. This requires to establish an accurate mathematical model of the heating furnace. But due to the tubular furnace is a kind of complicated multivariable system,it has the process characteristics of long time-delay,many interference factors,strong coupling,closed loop in the system and so on. If using the original subspace method there will be the problem of biased estimation. This paper uses an improved subspace identification algorithm by designing instrumental variables. The method can effectively solve the problem of biased identification results. By applying this method to a simulation example and the identification of the tubular furnace model in one enterprise,the simulation results show that the improved method has higher identification accuracy than the original subspace method. The improved method is more suitable for complex industrial environment.
出处
《计算机仿真》
CSCD
北大核心
2016年第2期273-276,294,共5页
Computer Simulation
基金
天津市科技支撑重点项目(10ZCKFGX03000)
科技部中小企业创新基金(09C262112022954)
天津市自然科学基金重点项目(07JCZDJC09600)
天津市高等学校科技发展基金计划项目(20120705)
关键词
子空间辨识
闭环辨识
变量有误差模型
辅助变量法
Subspace identification
Close-loop identification
Errors-in-variables models
Instrumental variable method
