高效再生催化裂化装置多稳态分析:反应温度开/闭环控制条件对热反馈机制的影响

针对催化裂化反应-再生系统在提升管反应温度开环和闭环控制条件下的输出与输入多稳态问题,分析了烧焦罐式高效再生催化裂化反应-再生系统在两种条件下随着CO助燃剂添加量变化时的多稳态分布。在反应温度开环条件下,因再生温度与反应温度的耦合程度较低,使系统移热曲线呈单调递增,导致了系统出现3个稳态操作点。在反应温度闭环控制条件下,提升管反应器和再生器间热反馈机制发生改变,由于再生剂循环量可以作为额外的自由度对再生温度和反应温度之差进行补偿,再生器和提升管反应器的耦合程度增强,使得系统只会在助燃剂添加量极低时才会出现多个稳态点,而在基准操作条件下只有一个稳态点,规避了系统在提升管反应温度开环时的多个稳态点的问题。     

CO燃烧状况可调的催化裂化装置多稳态分析

考虑到一氧化碳(CO)助燃剂作为调控系统CO燃烧状况的操作手段会影响催化裂化装置的多稳态,在已有的烧焦罐式高效再生催化裂化装置反应-再生系统动态机理模型的基础上,扩展了再生器内CO燃烧的机理模型,并将CO助燃剂的添加量与再生器内CO的燃烧状态进行定量的关联。在对再生器进行热量衡算的基础上,绘制出再生器的放出-移出热量曲线,确定了系统的3个稳态操作点,并根据放热反应器稳态热稳定性斜率条件分析了稳态点的热稳定性。针对炼厂CO助燃剂添加量设计较为保守的问题,分析了减少助燃剂添加量对系统操作的影响,指出CO助燃剂的添加量会影响到系统两个高温稳态点之间的操作空间从而影响到装置控制系统设计的苛刻度。     

催化裂化装置反应——再生系统的技术改造

系统地总结了３５ｋｔ／ａ催化裂化装置从ＲＦＣＣ工艺改造到７０ｋｔ／ａＡＲＧＧ工艺的反应－再生系统技术改造的经验,通过改造使催化装置不仅在工艺技术上产生了质的飞跃,而且在处理量上有显著的增加。     

进料温度对整个重油催化裂化装置反应的影响

对于无取热设施的重油催化裂化装置,反应进料温度对反—再系统热平衡有一定的调节作用,但是靠降低反应进料温度来维持反—再系统热平衡有一定的局限。由于雾化效果变差和剂油比增加导致焦炭产率增加,烧焦放热将会超过由反应进料温度降低而少带入反应器的热量,因此降低反应进料温度有一极限,超过该值对热平衡的调节作用就会发生逆转。     

前置烧焦式催化裂化装置稳态模拟:再生器取热器负荷对过程的影响

应用本文作者课题组开发的前置烧焦式催化裂化装置过程模拟平台,对某工业装置进行现场条件下的稳态模拟,并完成全装置模型的校正及验证。在此基础上,模拟分析了再生器取热器负荷对该装置操作产生的影响。模拟结果显示,改变再生器取热器的负荷,能够有效调节两器热量平衡,在确保工艺范围不超限的前提下提高装置剂油比,原料油转化率提高,装置产品分布得到有效调节,但是使得密相床再生器密相区床层温度降低,从而影响再生器的再生效果,不利于提升管进口再生催化剂初始活性的提高。     

反应条件对MIP催化裂化反应结果的影响

对中石化股份有限公司安庆分公司MIP催化裂化装置扩能改造前后两种工况对比,分析了在两种工况下,MIP催化裂化第二反应区发生的不同类型的氢转移反应,以及因此而导致的不同反应结果.反应结果表明,提高催化裂化平衡催化剂活性是降低汽油烯烃含量的有效方法,而且随着平衡催化剂活性的提高,宜适当提高第二反应区反应温度,这样会使第二反应区Ⅰ类型氢转移反应占优,可在保障降低汽油烯烃含量的基础上,获得较高的转化率和较好的产品分布.     

结焦对催化裂化装置长周期影响的分析

催化裂化装置关键设备的结焦严重制约了装置的长周期运行,如何通过生产操作和技术进步减少结焦是催化裂化装置必须面对的问题.以抚顺石油一厂催化裂化装置为例,分析论述了影响及抑制装置结焦的因素,从而为催化装置的长周期运行提供一些办法.     

基于动态优化的催化裂化装置再生器裕量分析与控制设计(Ⅱ) 求解方法与结果分析

用动态优化的方法求解催化裂化装置再生器的工艺裕量与控制设计。对于该多目标混合整数动态优化问题,通过ε-约束法处理多目标优化问题,将控制性能目标函数转化为控制性能约束;通过将0-1变量松弛化、引入附加等式约束求解混合整数动态优化问题。求解得到非劣解集,绘制关于控制性能指标和工艺指标的关系曲线,发现系统对控制器性能的要求愈高,所需要的裕量应愈大。由此进行折中处理,从而找到兼顾工艺要求和控制性能的催化裂化装置再生器优化设计方案。     

基于稳定性的循环物流系统流程模拟——以催化裂化反应-再生系统为例

过程流程模拟中广泛应用的序贯模块法处理循环物流系统存在很多困难,如断裂流股的选择、迭代方程的收敛性等。基于稳定性理论解决循环物流系统的流程模拟问题,首先根据化工单元装置的模型方程将其分为正向模型和反向模型,并将循环物流中的变量定义为迭代变量和收敛变量;然后在收敛变量处将循环物流的流股断裂,迭代变量分别通过正向模型和反向模型计算收敛变量,二者的偏差通过增益系数对迭代变量进行修正,进而得到迭代方程;最后,利用控制理论中的稳定性理论来确定迭代方程的增益系数,将迭代方程线性化,采用劳斯判据确定增益系数的稳定范围,当增益系数位于稳定范围以内时,迭代方程必然收敛。催化裂化装置反应-再生系统因催化剂循环的存在而成为典型的循环物流系统,本文将反应器作为正向模型,再生器作为反向模型,以再生温度和再生催化剂含碳量作为迭代变量,构造了催化裂化装置反应-再生系统流程模拟的迭代方程,利用稳定性理论找到增益系数的稳定范围,保证流程模拟计算必定达到收敛,验证了该方法的可行性和有效性。     

基于超熵产生分析的反应精馏系统多稳态研究(Ⅱ)稳定性分析

以环氧乙烷与水合成乙二醇为模型体系进行了实例分析,得到了该反应精馏系统在一定条件下的多定态解,利用超熵产生判据式对各定态解的稳定性进行了分析,在此基础上进一步讨论了多稳态产生的原因.结果表明:浓度涨落、温度涨落、其他操作负荷的变化以及反应动力学特性都可以成为反应精馏系统出现多稳态的原因.将超熵产生判据用于反应精馏体系定态稳定性的判定,较为可靠.因为超熵产生判据式是由热力学基本原理推导而来,较好地反映了反应精馏过程演化的非线性不可逆的热力学本质,因此本方法对反应精馏系统稳定性的研究,具有普遍的理论指导意义.     

The determination of possibility of multiple steady states for complex catalytic reactions in an isothermal CSTR

A class of complex reaction networks with deficiency from two to six, involving isothermal catalytic reactions in a continuous flow stirred tank reactor (CSTR), is determined to have the capacity to admit multiple positive steady states by implementation of the combination of the Deficiency One Algorithm and the Subnetwork Corollary. A set of rate constants and two corresponding positive steady states, one stable and the other one unstable, are constructed for four examples.     

The effect of non‐ideal mixing on the number of steady states and dynamic behaviour for autocatalytical reactions in a CSTR

In this paper, we analyze the concentration multiplicity and dynamic behavior for an autocatalytical reaction, A + R → (n + 1)R + products with an overall rate expression given by – γ_a= kc_a^pc_r^r(p > 0 and r > 0) in a imperfectly mixed (Cholette's model) CSTR. We proved that non‐ideal mixing had an effect on the number of steady states and dynamic behavior for the reaction orders r > 1 and r = 1. However, the above‐mentioned effect does not happen for the reaction order r < 1. Furthermore, a simulated example was used to demonstrate our results.     

一种新的生产计划与催化裂化装置过程操作集成的闭环策略

Production planning models generated by common modeling systems do not involve constraints for process operations, and a solution optimized by these models is called a quasi-optimal plan. The quasi-optimal plan cannot be executed in practice some time for no corresponding operating conditions. In order to determine a practi- cally feasible optimal plan and corresponding operating conditions of fluidized catalytic cracking unit （FCCU）, a novel close-loop integrated strategy, including determination of a quasi-optimal plan, search of operating conditions of FCCU and revision of the production planning model, was proposed in this article. In the strategy, a generalized genetic algorithm （GA） coupled with a sequential process simulator of FCCU was applied to search operating conditions implementing the quasi-optimal plan of FCCU and output the optimal individual in the GA search as a final genetic individual. When no corresponding operating conditions were found, the final genetic individual based correction （FGIC） method was presented to revise the production planning model, and then a new quasi-optimal production plan was determined. The above steps were repeated until a practically feasible optimal plan and corresponding operating conditions of FCCU were obtained. The close-loop integrated strategy was validated by two cases, and it was indicated that the strategy was efficient in determining a practically executed optimal plan and corresponding operating conditions of FCCU.     

An Extensive Analysis on the Start‐up of a Simple Distillation Column with Multiple Steady States

The impact of start‐up procedures on the behaviour of simple distillation columns with multiple steady states is analyzed. A well‐known system with hysteresis is studied in this paper and transient responses for a binary distillation column yielding different steady states during the start‐up operation are shown. Several dynamic simulation results showing an interesting behaviour are presented. Through the examination of the profile evolutions corresponding to given start‐up policies, it can be seen how the column arrives to different steady states. It is shown that it is possible to identify a set of critical values for the start‐up supervision. Also, guidelines of general validity are achieved with the aim of finding the appropriate start‐up policy to obtain the desired solution.     

Theoretical analysis of inhibition effect on steady states of enzymatic membrane reactor with substrate and product retention for reaction producing weak acid

The aim of this study is to formulate a model of enzymatic membrane reactor (EMR), i.e., a continuous, stirred tank bioreactor with full enzyme recycle, for a reaction producing a weak acid, and to explore the effect of substrate and product inhibition of different mechanisms coupled with transport properties of the membrane on the static behaviour of the system. The inhibition of an enzyme by a substrate leads to the non-monotonicity of reaction rate expression with respect to the substrate concentration. If a product of enzymatic reaction, taking place in the EMR, influences the pH of a reaction mixture this is also the factor causing the non-monotonicity of the substrate and product dependent reaction rate. The character of these dependencies affects substantially the structure of the steady states of the reactor. The bifurcation diagrams, shown in the work, are of different characters depending on the bifurcation parameter. It has been found, that bifurcation diagrams for competitive and uncompetitive inhibition by a substrate differ in the number and position of bifurcation points. Steady states of multiplicity five have been localised in case of uncompetitive inhibition by the substrate at high affinity of the enzyme to this substrate. Retention of reagents, related to transport properties of a membrane, influences significantly the effectiveness of a process. A specially written software in Delphi™ has been used for the calculations.     

Output feedback control of dissipative PDE systems with partial sensor information based on adaptive model reduction

We address the problem of control of spatially distributed processes in the presence of measurement constraints. Specifically, we assume the availability of sensors that measure part of the state spatial profile. The measurements are utilized for the derivation and on‐demand update of reduced order models (ROM) based on an extension of the adaptive proper orthogonal decomposition (APOD) method using a snapshot reconstruction technique. The proposed Gappy‐APOD methodology constructs locally accurate low‐dimensional ROM thus resulting in a computationally efficient alternative to using a large‐dimensional ROM with global validity. Based on the low‐dimensional ROM and continuous measurements available from point sensors a Lyapunov‐based static output feedback controller is subsequently designed. The proposed controller design method is illustrated on an unstable process modeled by the Kuramoto‐Sivashinsky equation, when the designed controller successfully stabilizes the process even in the presence of model uncertainty. © 2012 American Institute of Chemical Engineers AIChE J, 59: 747–760, 2013     

基于高斯模型的聚乙烯过程设计与控制集成优化

聚乙烯反应过程中物流-能流剧烈交叠、反应-传递相互耦合,使得过程具有强非线性以及多重稳态。传统的顺序设计方法不能保证系统有足够的控制自由度,当存在扰动和过程参数不确定性时,仅依靠设计控制器很难提高产品质量。提出一种聚乙烯工艺稳态设计与运行控制的集成优化方案,创造性地引入Kriging高斯模型同时预测模型动态和模型不确定性。另一个重要的贡献是在聚乙烯工艺设计阶段,设计性能指标,定量描述过程稳态设计对闭环动态的影响。所提出的方法已经通过对气相聚乙烯工艺设计和运行控制的集成优化进行了验证,并在参数不确定性和扰动存在情况下仿真证实了集成优化设计方案的高效性。     

Multi-agent reinforcement learning for process control: Exploring the intersection between fields of reinforcement learning,control theory,and game theory

The application of reinforcement learning (RL) in process control has garnered increasing research attention. However, much of the current literature is focused on training and deploying a single RL agent. The application of multi-agent reinforcement learning (MARL) has not been fully explored in process control. This work aims to: (i) develop a unique RL agent configuration that is suitable in a MARL control system for multiloop control, (ii) demonstrate the efficacy of MARL systems in controlling multiloop process that even exhibit strong interactions, and (iii) conduct a comparative study of the performance of MARL systems trained with different game-theoretic strategies. First, we propose a design of an RL agent configuration that combines the functionalities of a feedback controller and a decoupler in a control loop. Thereafter, we deploy two such agents to form a MARL system that learns how to control a two-input, two-output system that exhibits strong interactions. After training, the MARL system shows effective control performance on the process. With further simulations, we examine how the MARL control system performs with increasing levels of process interaction and when trained with reward function configurations based on different game-theoretic strategies (i.e., pure cooperation and mixed strategies). The results show that the performance of the MARL system is weakly dependent on the reward function configuration for systems with weak to moderate loop interactions. The MARL system with mixed strategies appears to perform marginally better than MARL under pure cooperation in systems with very strong loop interactions.