节点文献

精馏节能过程的wave非线性动态建模与非线性控制策略研究

Wave Based Nonlinear Dynamic Modeling and Nonlinear Control Strategy Design for Energy-saving Distillation Processes

【作者】 周叶翔

【导师】 刘兴高;

【作者基本信息】 浙江大学 , 系统工程, 2010, 硕士

【摘要】 精馏节能控制一直是国际节能控制研究的重点和热点。内部热耦合精馏塔节能潜力可达常规精馏塔的30%以上,具有极大的应有前景。由于内部热耦合精馏过程具有强耦合、强病态、强不对称性、强反向响应等复杂非线性动态特性,其控制设计一直是阻碍该高效节能技术商业化的瓶颈问题。目前常规控制方案如PID,内模控制等即使能够控稳热耦精馏节能过程,但控制质量不好,特别是在高纯时目前已报道的各种控制方案难以适用,究其原因可能在于近似线性模型、基于数据辨识的统计模型等不能有效描述热耦合精馏过程的复杂动态特性,导致基于传统模型的各种先控方案对于控制效果的改善也有限。因此本文首先从内部热耦合精馏过程的非线性动态建模工作做起,然后建立内部热耦合精馏塔和内部热耦合空分塔的非线性控制策略。论文主要工作与贡献如下:1.首次对内部热耦合精馏塔(ITCDIC)的Wave传播现象进行研究,推导出内部热耦合精馏塔的非恒摩尔流自然波速、激波波速、波形描述函数等。结合热耦合关系,建立内部热耦合精馏塔的完整Wave非线性动态模型。在Wave分析的基础上,首次分析研究内部热耦合精馏塔的干扰脉冲传播特性,并综合分析内部热耦合精馏塔的不对称动态特性、反向响应等独特动态特性。2.基于内部热耦合精馏塔Wave非线性动态模型以及Wave传播特性,建立了三种一般模型控制、两种广义一般模型控制策略。五种控制策略有效地克服了内部热耦合精馏塔各种传统控制方案体现出来的控制品质不高、鲁棒性不强等弊端。五种控制方案都结合了Wave模型,很好地解决了控制变量之间的耦合关联问题,实现了自然解耦。其中,基于振荡波速的改进广义一般模型控制方案在不牺牲塔底产品纯度的情况下,突破性地将塔顶纯度控制设定值提高到超高纯领域(99.999%)。3.首次研究了内部热耦合空分塔(ITCASC)的Wave传播现象,推导出了内部热耦合空分塔的非恒摩尔流自然波速、激波波速公式。并以液氮液氧的浓度波形为例,分析了内部热耦合空分塔的波形相干性、干扰脉冲传播等特征。同时,结合Wave理论,对内部热耦合空分塔的非线性动态特性进行了研究,首次发现了内部热耦合空分塔强烈的反向响应等动态特性。4.基于内部热耦合空分塔的开环响应特性研究,设计出了两种不同控制变量组合下的PI控制方案,并进一步设计出了有效的ITCASC一般模型控制方案和非线性模型预测控制方案。研究结果表明,一般模型控制方案因为模型失配问题,干扰控制效果反而较PI控制效果差;两种PI控制方案的对比证明采用侧提流量作为控制变量,虽然能减弱控制回路之间的耦合关系,但是对塔顶控制效果不是很好;而非线性模型预测控制方案有效地解决了模型失配的问题,并避免了控制解耦问题,在设定值控制、干扰控制方面都体现了不错的效果。

【Abstract】 Energy saving technology has become a key point in the energy saving research area. Internal thermally coupled distillation has very high energy saving potential. Compared with conventional distillation columns it could save more than 30% energy. However, the complex nonlinear dynamic performance of internal thermally coupled distillation processes brings a great challenge to the control design. Traditional control strategies, like PID, internal model control and so on, could hardly control internal thermally coupled distillation processes, the control performance need more improvement. The strong asymmetric character, inverse response dynamic and ill condition performance make the efficiency of those ITCDiC control strategies reported in literature decreased seriously under high-purity control conditions, which is mainly because of the model mismatch of those models. As a result, many advanced control strategies could hardly improve the control performance. In the present work, Wave theory is exploited to low-order nonlinear modeling and nonlinear controller design of ITCDIC and ITCASC.The main work and contributions are summarized as follows:1. Pioneered in the Wave traveling disturbance pulse traveling analysis of ITCDIC, the development of natural wave velocity, shock wave velocity with varied molar flow rate and profile description function are carried out; a complete nonlinear wave dynamical model is then established; based on wave traveling analysis, the nonlinear dynamic performance like asymmetric and inverse response character are futher explored;2. Based on wave model of ITCDIC, three generic model control strategies, two improved general generic model control strategies are proposed, which well improve the model mismatch problem and robust performance, realize the multivariable decoupling function automatically. Specially, the proposed XGGMC control strategy realize the super high-purity control of ITCDIC with 99.999% product purity;3. The wave traveling phenomena and disturbance pulse traveling in ITCASC are explored, the natural wave velocity, shock wave velocity of ITCASC with varied molar flow rate are then developed, combined with which, the distinct nonlinear dynamic behavior like asymmetric and inverse response performance are explored for the first time;4. Based on the open-loop response analysis, PI control strategies, generic control strategy (GMC) and nonlinear model predictive control strategy (NMPC) of ITCASC are futher established. Detailed comparison research results show that NMPC present the best control performance, the control quality of generic model control deteriorate seriously in the regulatory control because of the model mismatch of the polynomial model. The two PI control strategies show that employing side extraction flow rate as a control variable could decrease the coupling relationship between the control loops efficiently, however the control quality of the Nitrogen concentration deteriorates seriously compared with the PI controller employing the pressure of rectifying section as a control variable.

  • 【网络出版投稿人】 浙江大学
  • 【网络出版年期】2010年 08期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络