高级检索

    连续搅拌釜的建模与模型预测温度控制方法

    Modeling and model predictive temperature control of a CSTR system

    • 摘要: 连续搅拌反应釜(CSTR)温度控制是化工过程中的核心环节,其精度直接影响产品质量与系统能耗。传统串级比例-积分-微分(PID)控制因其在跟踪性能和抑制扰动等方面能力不足,难以满足复杂工况的跟踪和抗干扰需求,文中基于反应釜传热与釜内反应热力学模型,搭建了更符合实际工业场景的控制系统模型。结合CSTR内环冷却的特点,提出一种改进的无偏模型预测控制(MPC)方法。通过建立反应釜内环模型的输入扰动扩展模型,运用卡尔曼滤波器,实现状态变量与扰动的同时估计。对比实验的结果表明,相较于传统串级PID,文中提出的无偏MPC方法在温度跟踪性和系统鲁棒性方面有明显改善。研究结论为实际扰动环境下反应釜温度的控制提供了参考。

       

      Abstract: Temperature control in a CSTR influences product quality and energy consumption. Cascade PID is difficult to meet tracking accuracy and disturbance rejection requirements under complex working conditions due to its deficient tracking and disturbance rejection ability. This study built a control system based on the heat transfer and reaction models, which could better match actual industrial models. By establishing an extended inner-loop model combined with the inner-loop cooling characteristics of the CSTR, an improved offset-free MPC method was proposed with Kalman filter for estimating the states and disturbances. Contrast experiments showed that the proposed Offset-free MPC method exhibited obvious improvement in temperature tracking accuracy and robustness compared to cascade PID. The conclusion of this paper provides a reference for temperature control of a CSTR under actual disturbance environment.

       

    /

    返回文章
    返回