基于分阶段的LSSVM发酵过程建模

发酵过程建模是研究微生物发酵的重要课题,基于模型可实现被测参量的软测量、系统的优化控制。鉴于引入混合核函数的最小二乘支持向量机在过程建模中具有优良表现,采用基于混合核函数的最小二乘支持向量机建模。但由于发酵过程周期较长,最小二乘支持向量机的全局模型预测精度难以保证,算法复杂度很高,因此提出一种分阶段建模方法。首先,选择表征阶段特性的辅助变量,利用模糊C均值聚类算法对样本数据聚类,将发酵过程分成不同的阶段,然后为各个阶段分别建立最优混合核最小二乘支持向量机局部模型,最后将局部模型合成构成过程的完整模型。将此方法应用于青霉素发酵过程和重组大肠杆菌发酵过程中,验证了该方法的有效性。     

发酵过程生物量软测量技术的研究进展

生物量是发酵过程中的关键过程参数之一,它直接影响着发酵过程的优化和控制。综述了近年来发酵过程生物量软测量技术的研究现状,讨论了基于过程机理分析、回归分析、状态估计和神经网络等的软测量建模方法,对基于神经网络和改进的神经网络建模方法进行了分析。指出基于多尺度建立软测量混合模型,是实现发酵过程生物量在线测量的有效方法,并给出了建立混合模型需要解决的关键问题。     

一种提高微生物发酵单位的新优化控制策略

为了提高微生物发酵单位,提出了支持向量机（SVM）与基于实数编码遗传算法（RGA）相互耦联的优化控制策略。为解决发酵参数之间的耦合问题,进一步引入了模式的概念。SVM建立微生物发酵过程的预估模型,RGA以此模型为适应度函数计算最优控制模式。此策略应用于青霉素发酵过程的优化控制,效价比没有实施优化控制策略的生产结果提高了22.88％。     

青霉素发酵过程中的混合建模

由于微生物发酵过程机理的复杂性和高度非线性,建立发酵过程的精确模型具有一定难度。传统的动力学模型预测虽然会与实际输出有一定偏差,但它在某种程度上反映了过程机理;神经网络建模方法属于"黑箱"方法,建模过程中没有用到任何先验知识,有一定的预测效果;因此单一的建模方法往往会不具备其他建模方法的优势。本文以传统的发酵动力学模型为基础,结合RBF神经网络进行混合建模的"灰箱"建模方法是一种比较好的建模思路,可以取得较满意的软测量效果。     

基于pH控制的丙酮丁醇间歇发酵过程动力学模型

基于微生物发酵动力学模型和丙酮丁醇发酵工艺的特点,由5组控制pH下的实验数据建立该发酵过程的动力学模型,采用改进的自适应遗传算法对模型的12个参数进行优化,并采用一组控制pH数据对模型进行了外推验证,结果表明,该模型能够较好地描述pH的丙酮丁醇间歇发酵过程.     

基于LS-SVM和Pensim仿真平台的青霉素发酵过程建模

从青霉素发酵过程仿真平台(Pensim)得到的结果作为出发点,采用最小二乘支持向量机(LS-SVM)对青霉素发酵过程进行建模研究。分别研究丁利用溶解氧浓度、排气二氧化碳浓度等变量对青霉素产物浓度、菌体浓度和底物浓度等重要过程变量的建模问题,在3种不同的仿真条件下分别建立了相应的在线预报模型,并对其进行了分析和比较。基于 Pensim 的仿真结果表明采用 LS-SVM 方法所建立的在线预报模型均具有良好的预测精度,对后续发酵过程的控制和优化能起到一定的参考作用。     

基于模糊满意度的烧结过程多目标优化控制

为实现铁矿石烧结过程中烧结终点和混合料槽料位两个关键参数的综合优化,提出了一种基于模糊满意度的多目标优化控制方法。首先,通过综合运用模糊控制、预测控制的切换控制技术,建立烧结终点的智能优化控制模型;其次,为稳定混合料槽料位,基于专家知识,建立了混合料槽料位专家控制模型;最后通过构造求解多目标优化问题的系统满意度函数,将多目标优化问题转化为单目标优化问题,求取系统的满意解,实现烧结过程多目标优化控制。仿真结果和实际运行结果表明提出的方法是可行而有效的。     

改进粒子群算法在酶发酵过程优化控制中的应用

针对发酵过程的非线性、时变等特点,建立了基于BP神经网络的产物浓度预估模型,并在此模型的基础上,利用改进的粒子群算法(PSO)实现对L-天冬酰胺酶Ⅱ发酵过程控制参数的寻优,以确定该过程的优化轨线,通过在线调整实现对优化轨线的跟踪控制,实验应用结果表明效果良好.     

炼焦生产过程质量产量能耗的集成优化控制

针对炼焦生产过程强非线性、大时滞等特点,基于过程参数的主元分析和灰色关联分析,建立了焦炭质量、产量及焦炉能耗的神经网络预测模型和以焦炭质量为约束条件,产量、能耗为目标函数的优化控制模型。提出一种融合模糊C均值聚类粗优化和差分进化细优化的集成优化控制方法,进行过程参数的优化并给出操作优化指导。系统仿真结果表明,该方法能有效地抑制工况的波动,达到高产、优质、低耗的生产目标,为复杂工业过程的建模和优化控制提供了一种新思路。     

抗生素发酵过程优化调度模型的研究

对抗生素多罐并行发酵过程进行了分析,将任务、设备和事件之间的分配关系表达为两类0-1变量,建立了一个基于连续时间的抗生素多罐并行发酵过程优化调度的M ILP(混合整数线性规划)模型。该模型整数变量少,求解速度快,并给出了最短生产时间和罐批最优生产序列。最后以头孢菌素发酵过程为例证明了此调度模型的可行性和有效性。     

Suboptimal control through model reduction

There has been a growing interest in control strategies that optimize a given performance criterion to improve process efficiencies. Optimal control techniques, such as LQP, have been available but did not find much acceptance in process industries because of certain problems that are involved in their application. In this paper an approach is presented which alleviates these problems. Simulation results show that the proposed approach works well for servo and regulatory control problems.     

A feedback control framework for safe and economically‐optimal operation of nonlinear processes

Maintaining safe operation of chemical processes and meeting environmental constraints are issues of paramount importance in the area of process systems and control engineering, and are ideally achieved while maximizing economic profit. It has long been argued that process safety is fundamentally a process control problem, yet few research efforts have been directed toward integrating the rather disparate domains of process safety and process control. Economic model predictive control (EMPC) has attracted significant attention recently due to its ability to optimize process operation accounting directly for process economics considerations. However, there is very limited work on the problem of integrating safety considerations in EMPC to ensure simultaneous safe operation and maximization of process profit. Motivated by the above considerations, this work develops three EMPC schemes that adjust in real‐time the size of the safety sets in which the process state should reside to ensure safe process operation and feedback control of the process state while optimizing economics via time‐varying process operation. Recursive feasibility and closed‐loop stability are established for a sufficiently small EMPC sampling period. The proposed schemes, which effectively integrate feedback control, process economics, and safety considerations, are demonstrated with a chemical process example. © 2016 American Institute of Chemical Engineers AIChE J, 62: 2391–2409, 2016     

Graph reduction of complex energy‐integrated networks: Process systems applications

We illustrate the application of a graph reduction method developed recently to analyze complex energy‐integrated process networks. The method uses information on the energy flow structure of the network and the orders of magnitude of the different energy flows to generate, automatically, information on the time scales where the process units evolve, canonical forms of the reduced models in each time scale, and controlled variables and potential manipulated inputs available in each time scale. Representative examples of reactor‐heat exchanger and distillation column networks are considered to illustrate the method and develop insights on effective control strategies for these processes. © 2014 American Institute of Chemical Engineers AIChE J, 60: 995–1012, 2014     

Nonlinear PI controllers with output transformations

Well‐designed nonlinear proportional‐integral (PI) controllers are successful for nonlinear dynamical processes like linear PI controllers are for linear processes. Two nonlinear blocks representing proportional and integral terms can be designed so that the linearized controllers perform the same as linear PI controllers for linearized processes at the given operating points. For some nonlinear processes, nonlinear blocks for nonlinear PI controllers can be singular at some operating points, and control performances can be poor for set points near those points. To mitigate such disadvantages, new nonlinear PI controllers that introduce output transformations are proposed. Several examples are given, showing the performance of the proposed nonlinear PI controllers. © 2015 American Institute of Chemical Engineers AIChE J, 61: 4264–4269, 2015     

化工过程非稳态开工的缓冲升温修正切换控制

化工过程的开工大多表现为一个升温过程。为满足升温过程的快速性而又不失稳定性的要求,有学者提出将时间最优Bang-Bang控制与其他控制方法结合来控制升温过程,但由于Bang-Bang控制对切换次数和切换点要求严格,致使其在实际应用中不够理想。在Bang-Bang组合控制的基础上引入缓冲升温控制,将整个控制系统分为4部分:全幅升温、全幅降温、缓冲升温、PID控制。将温度变化率作为缓冲升温与PID控制的切换变量,将修正切换控制问题等价为非线性规划问题,优化选取全幅升温、全幅降温、缓冲升温最优切换时间点。实例对比分析表明,该切换开工控制方案不仅避免了因切换次数与切换点选取不当导致的不良问题,且超调小,稳定快。     

统计过程控制(SPC)在精密管材生产中的应用

简要介绍统计过程控制的原理,总结影响管材挤出流率的因素,并通过精密管材SPC实验结果进行了分析。     

过程控制系统在热处理炉的应用

介绍莱铜热处理炉过程控制系统，其基于拥有自主知识产权的过程控制开发平台，采用有限差分模型预报炉内铜板加热过程，实现了热处理计划管理、工艺参数设定、钢板温度跟踪、加热过程动态控制、生产数据统计等功能。     

Robust model predictive control and fault handling of batch processes

This work considers the control of batch processes subject to input constraints and model uncertainty with the objective of achieving a desired product quality. First, a computationally efficient nonlinear robust Model Predictive Control (MPC) is designed. The robust MPC scheme uses robust reverse‐time reachability regions (RTRRs), which we define as the set of process states that can be driven to a desired neighborhood of the target end‐point subject to input constraints and model uncertainty. A multilevel optimization‐based algorithm to generate robust RTRRs for specified uncertainty bounds is presented. We then consider the problem of uncertain batch processes subject to finite duration faults in the control actuators. Using the robust RTRR‐based MPC as the main tool, a robust safe‐steering framework is developed to address the problem of how to operate the functioning inputs during the fault repair period to ensure that the desired end‐point neighborhood can be reached upon recovery of the full control effort. The applicability of the proposed robust RTRR‐based controller and safe‐steering framework subject to limited availability of measurements and sensor noise are illustrated using a fed‐batch reactor system. © 2010 American Institute of Chemical Engineers AIChE J, 2011     

Design and control of the ethyl benzene process

The ethyl benzene (EB) process involves the reaction of benzene with ethylene to form the desired EB product. However, ethylene can also react with EB to form an undesired product of di‐ethyl benzene (DEB) if reactor temperatures or ethylene concentrations are high. An unusual feature of the EB process is the ability to recycle “to extinction” all the DEB formed in the reactor (no net DEB product produced), since DEB reacts with benzene to form EB. Since DEB is the highest‐boiling component in the system, it comes out the bottom of the two distillation columns, so there is little energy penalty in having a large DEB recycle. Recycling benzene is more expensive because it goes overhead in the first distillation column. The economic optimum steady‐state design is developed that minimizes total annual cost (capital and energy). An effective plantwide control structure is also developed. © 2010 American Institute of Chemical Engineers AIChE J, 2011