MPCE实验装置列管式换热器系统建模与控制算法研究

【作者】 钟柳芳；

【导师】 喻寿益；

【作者基本信息】 中南大学 ， 控制科学与工程， 2010， 硕士

【摘要】 换热器是一种用来进行热量交换的工艺设备,被广泛应用于流程工业生产企业。在目前尚不具备在实验室中复现真实工业过程的条件下,研究利用具有典型对象模型的实验装置进行模拟仿真无疑是一条探索将控制理论和算法转化为应用技术的捷径。以多功能过程控制实验装置(MPCE)为实验平台,以列管式换热器系统为被控对象,对其进行建模、控制算法以及换热过程模拟仿真的研究,对于改善换热器系统的控制效果,提高换热效率具有重要的意义。利用机理建模与系统辨识的方法建立了换热器的数学模型,确定了换热器热流出口温度与冷流入口流量之间的关系。通过阶跃响应仿真实验和采用PID控制实验验证了模型的准确性。并利用较准确的换热器模型,在MPCE实验装置上实现了Smith预估控制实验和前馈—反馈控制实验。针对换热器大时滞、参数时变的特性,研究了全参数自适应Smith预估控制算法(APASPCA).在常规Smith预估器的基础上,基于模型参考自适应原理,利用Lyapunov函数法设计了预估器增益和时间常数的自适应律,并利用信号相关性分析技术中的平均幅度差函数实现了被控对象纯滞后时间的最优估计,使得预估器模型参数能跟踪被控对象参数的变化,从而保证了Smith预估器的控制效果。仿真结果表明,APASPCA算法对于大时滞参数时变的被控对象具有很好的控制效果,是一种性能很好的控制方法,具有实际应用价值。更多还原

【Abstract】 Heat exchanger is a kind of process equipment for heat exchange which is widely used in process industry manufacturers. Under the condition that cannot recover real industrial processes, researching on typical experiment equipment with typical object model undoubtedly is a shortcut to explore control theory and algorithm into application technology. Using the multifunction process and control experiment (MPCE) device as experiment platform and tube heat exchanger as the controlled object, researches on the modeling and control algorithm of tube heat exchanger and simulation of heat transfer process are significant to improve control effect of heat exchanger system and enhance exchange thermal efficiency.Mechanism modeling and system identification methods were adopted to develop a mathematical model of heat exchanger, and determined the relationship between outlet temperature of heat fluid and inlet flow of cold fluid of heat exchanger. Step response simulation experiments and PID control experiments were conducted to verify the accuracy of the model.Based on the relatively accurate heat exchanger model,Smith predictive control experiments and feedforward-feedback experiments were realized on MPCE device.Aiming at the heat exchanger with large time delay and time-varying characteristics, all parameters adaptive Smith predictive control algorithm (APASPCA) was researched. Based on normal Smith predictor and model reference adaptive control principle, the adaptive laws of predictor gain and time constant were designed by adopting Lyapunov function method. Meanwhile the average magnitude difference function in correlation analysis technique was introduced to realize optimal estimation of time delay of controlled object which made real-time regulation of predictor parameters match with controlled object parameters. The algorithm could ensure the control effect of Smith predictor. Simulation results showed that APASPCA could get good control effect for the controlled object with large time delay and time-varying characteristics. The algorithm had good control performance and practical value.更多还原