基于动态模型的在线反馈优化

针对化工过程动态波动明显、优化模型存在较多的不确定性等特点,提出了一种考虑过程不确定性、基于过程动态模型的在线反馈优化策略。将过程动态模型按一定周期离散化为差分方程,基于差分方程进行动态优化,优化目标函数为优化时域的终端时刻的经济指标,优化变量为过程的操作变量,采用非线性规划作为优化算法;优化结果在实施后根据可测输出进行在线反馈,在优化模型的差分方程中引入误差修正项,将对应时刻的状态变量和相关变量的实际值代入可求出误差修正项,从而实现在线反馈优化。仿真结果表明,与传统的稳态操作优化相比,基于动态模型的反馈优化同样可将过程运行于最优操作点,同时具有很强的实时性,在外界干扰出现时可以立即作出反应,将过程推向最优操作点。     

局部最小二乘支持向量机回归在线建模方法及其在间歇过程的应用

当间歇生产切换于不同的工艺条件时,由于新工况下的样本一般很少,且批次间存在着不确定性（由于原材料波动或过程动态特性波动等）,基于全局学习的建模方法（如最小二乘支持向量机回归,LSSVR）建立的模型泛化性能不强。将局部学习融入LSSVR中,提出一种局部LSSVR（local LSSVR, LLSSVR）的间歇过程在线建模方法。结合前一批次离线优化后的LSSVR参数,针对待预测新样本在线选择与之相关的近邻样本集并基于此进行建模。以建立青霉素发酵过程的菌体浓度为例,验证了LLSSVR算法能够从过程的第2个生产批次开始在线建立较准确的预报模型,较LSSVR有着更好的推广能力、适应性和鲁棒性。     

操作时间不确定的间歇过程最优化设计(Ⅰ)不限制等待时间过程

对操作时间不确定的不限制等待时间的间歇过程的优化设计问题 ,提出一种在操作时间不确定条件下确定过程的限定循环时间的方法。该方法使模型的限定循环时间变为相互独立的 ,从而可将原过程转化为更新过程 ,并根据更新过程原理建立了优化该过程的期望值模型。算例表明考虑操作时间不确定性的优化结果要好于不考虑操作时间不确定性的优化结果。通过Monte-Carlo模拟表明采用新模型优化的结果在实际生产时是可行的     

间歇过程优化与先进控制综述

总结近年来间歇过程操作优化和设计优化中出现的各种新方法,以及在优化问题求解中使用的各种先进控制策略,反映间歇过程最优化和先进控制的最新研究方向。重点介绍间歇过程单元的操作优化和控制,兼顾在线稳态优化和动态优化。对新的研究方法提出展望。     

多因素不确定条件下的间歇生产调度优化

不确定条件下的间歇生产调度优化是生产调度问题研究中具有挑战性的课题。提出了一种基于混合整数线性规划(MILP)的鲁棒优化模型,来优化不确定条件下的生产调度决策。考虑到生产过程中的操作成本和原料成本,建立了以净利润最大为调度目标的确定性数学模型。然后考虑需求、处理时间、市场价格三种不确定因素,建立可调整保守程度的鲁棒优化模型并转换成鲁棒对应模型。实例结果表明,鲁棒优化的间歇生产调度模型较确定性模型利润减少,但生产任务数量增加,设备空闲时间缩短,从而增强了调度方案的可靠性,实现了不确定条件下生产操作性和经济性的综合优化。     

基于广义析取动态优化模型综合间歇精馏系统

基于数学规划技术提出间歇精馏过程设计与操作同步优化方法。首先,在原状态空间超级结构基础上嵌入时间维度,形成状态-时-空间超级结构,扩大间歇精馏过程优化空间。其次,将广义析取规划的思想引入间歇精馏动态优化模型,增加模型的直观性和可扩展性,并通过逻辑与析取约束,对方程进行合理分类,有效降低冗余方程导致的计算复杂度。新的优化策略采用严格模型,去除了恒摩尔流等简化假设,通过综合权衡设备投资、公用工程费用和操作周期等因素,一步获得最佳的间歇精馏塔型配置、结构参数和决策变量的动态控制方案。最后,以不同条件下苯-甲苯二元物系的间歇精馏分离为例展示了其有效性。     

基于长短时记忆神经网络和支持向量数据描述的间歇过程监测方法

间歇过程操作是化工过程中的一种重要生产方式.与连续过程不同,间歇生产不是在一个稳定的工作状态运行,而是根据设定的原料比例、操作条件所对应的操作轨迹运行.因此间歇过程数据具有多阶段性、动态时变性和非线性等特性,传统的监测方法难以应用于对间歇过程生产运行状态的监测.为了解决这个问题,提出了一种新的间歇过程监测策略.首先基于...     

不确定性间歇过程的一种实时优化控制方法

针对不确定性间歇过程的实时优化问题,提出了一种集成批间和批内优化的新实时优化控制方法。首先求解标称模型得到最优输入轨迹的结构,然后将输入轨迹参数化为若干标量决策变量及子输入轨迹。对最优性条件中的终端约束部分,使用批间优化满足约束条件;对梯度轨迹部分,提出一种回归法近似最优输入轨迹,以解决不确定扰动的在线不可测问题,实现梯度轨迹的批内优化。对一个间歇反应器的仿真研究表明了新方法的有效性。     

不确定性间歇过程的一种实时优化控制方法

针对不确定性间歇过程的实时优化问题,提出了一种集成批间和批内优化的新实时优化控制方法。首先求解标称模型得到最优输入轨迹的结构,然后将输入轨迹参数化为若干标量决策变量及子输入轨迹。对最优性条件中的终端约束部分,使用批间优化满足约束条件;对梯度轨迹部分,提出一种回归法近似最优输入轨迹,以解决不确定扰动的在线不可测问题,实现梯度轨迹的批内优化。对一个间歇反应器的仿真研究表明了新方法的有效性。     

基于MKPLS和SQP方法的间歇过程迭代优化控制

采用多向核偏最小二乘（MKPLS）方法建立间歇过程的模型并进行操作条件的优化。由于存在模型失配和未知扰动，基于MKPLS模型的最优控制轨迹在实际对象上往往难以实现最优的产品质量指标。本文利用间歇过程批次间的重复特性与序贯二次规划（SQP）优化算法中迭代计算的相似特点，提出了一种基于MKPLS模型的批次间优化调整策略，使得经过逐步优化调整得到的控制轨迹作用于实际对象时，可以得到更优的质量指标。该方法的有效性在苯乙烯聚合反应器和乙醇流加发酵过程的仿真对象上得到了验证。     

Nichtlineare stochastische Optimierung unter Unsicherheiten

Nonlinear Stochastic Optimization under Uncertainty Robust decision making under uncertainty is considered to be of fundamental importance in numerous disciplines and application areas. In dynamic chemical processes in particular there are parameters which are usually uncertain, but may have a large impact on equipment decisions, plant operability, and economic analysis. Thus the consideration of the stochastic property of the uncertainties in the optimization approach is necessary for robust process design and operation. As a part of it, efficient chance constrained programming has become an important field of research in process systems engineering. A new approach is presented and applied for stochastic optimization problems of batch distillation with a detailed dynamic process model.     

分批补料反应过程的非固定终端经济优化控制

针对批次生产周期不确定问题,提出一种非固定终端的经济优化控制方法。首先采用经济模型预测控制方法,用收益最大化的经济型目标函数代替终端约束,并将批次生产周期纳入被优化变量,建立动态经济优化问题,并通过对每个控制变量进行有差异的参数化,将动态优化问题转化为非线性规划（NLP）问题;然后使用内点罚函数法求解含非线性约束的优化问题,得到的最优控制序列和最佳批次生产周期,可将不确定扰动带来的损失降低到最小。其次采用非固定预测时域的滚动时域控制方法,不仅提高多变量系统的协同控制能力,而且根据实时预测终端产品产量不断优化更新关键操纵变量的控制分段函数的分割数及控制序列,从而可灵活优化操纵变量和操作时间的轨迹。最后在苯胺加氢过程上进行了批次优化控制性能测试,测试结果表明,非固定终端的经济优化控制从批次的总生产效益角度来优化每个批次生产的操作条件,实现批次反应过程生产时间与经济效益的最优化管理。     

基于样条插值模型的间歇精馏模拟与预测控制

以Aspen Batch Distillation（ABD）中的间歇精馏仿真系统为过程原型,提出了利用过程的模拟测试数据来建立间歇精馏过程的样条插值简化模型（spline interpolation model, SIM）。结合变回流比下的动态修正函数,构造出了一种简单实用的动态模型。该模型可有效模拟不同组分浓度下回流比发生变化时馏出液浓度和流量的动态变化情况。以该模型作为预测模型,进一步提出了一种变回流比的预测控制（model predictive control, MPC）算法来使馏出液浓度按照期望的设定值变化。控制仿真结果表明该控制方案计算简单,同时具有较好的控制效果。     

An in silico evaluation of data‐driven optimization of biopharmaceutical processes

Two methodological improvements of the design of dynamic experiments (C. Georgakis, Ind Eng Chem Res. 2013) for the modeling and optimization of (semi‐) batch processes are proposed. Their effectiveness is evaluated in two representative classes of biopharmaceutical processes. First, we incorporate prior process knowledge in the design of the experiments. Many batch processes and, in particular, biopharmaceutical processes are usually not understood completely to enable the development of an accurate knowledge‐driven model. However, partial process knowledge is often available and should not be ignored. We demonstrate here how to incorporate such knowledge. Second, we introduce an evolutionary modeling and optimization approach to minimize the initial number of experiments in the face of budgetary and time constraints. The proposed approach starts with the estimation of only a linear Response Surface Model, which requires the minimum number of experiments. Accounting for the model's uncertainty, the proposed approach calculates a process optimum that meets a maximum uncertainty constraint. © 2017 American Institute of Chemical Engineers AIChE J, 63: 2796–2805, 2017     

Nonlinear model of pneumatic conveying dryer for economic process control

Abstract

This paper presents a nonlinear dynamic model, suitable for economic process control of pneumatic conveying dryer for drying of food grains. The dynamic model is developed by reshaping the process equations derived for the batch drying, dilute phase, and a negative-pressure conveying system. The dynamic model parameters are identified by numerically solving a nonlinear least squares optimization problem, subject to a set of differential and algebraic equality constraints that describe the system dynamics and bounds in the parameters. A detailed parametric uncertainty and sensitivity analysis are performed providing valuable insight into the influence of critical model parameters on observables, the interplay among various parameter-state-measured disturbances, and quantifying uncertainties in the model. Further, different process economic performance and product quality indicator of uncertain dryer model are studied. The model validation study as performed with the underlying process shows a very good agreement in understanding necessary dynamic characteristics and interplay between the various parameter of interest.     

Use of tendency models and their uncertainty in the design of state estimators for batch reactors

Tendency models have been successful in the modeling and optimization of batch reactor processes where a detailed understanding based on fundamental principles and detailed kinetic studies is not available. The evolutionary nature of the Tendency modeling algorithm has proven useful in updating the process model between batches, as new process data or insight become available. But optimization is not the only task that can be undertaken with a Tendency model. In this work, the use of Tendency models in the design of state estimators to estimate reactor concentrations is investigated. The primary goal is to use the knowledge of the uncertainty in the Tendency model (which, by its nature, is an approximate model) to tune an extended Kalman filter. Two examples are presented to illustrate that even though Tendency models can feature a significant amount of uncertainty, they can be used successfully in state estimators.     

Dynamic optimization of a semi-batch reactor for polyurethane production

In this work, the dynamic optimization of a polyurethane copolymerization reactor is addressed. A kinetic-probabilistic model is used to describe the nonlinear step-growth polymerization of a mixture of low- and high-molecular-weight diols, and a low-molecular-weight diisocyanate. The dynamic optimization formulation gives rise to a highly complex and nonlinear differential-algebraic equation (DAE) system. The DAE optimization problem is solved using a simultaneous approach (SDO) wherein the differential and algebraic variables are fully discretized leading to a large-scale nonlinear programming (NLP) problem. The main reactor operation process control objective is the maximization of the molecular weight distribution (MWD) under a desired batch time, subject to a large set of operational constraints, while simultaneously avoiding the formation of polymer network (gel molecule). Typically, polyurethane formation is carried out using batch reactors. However, batch operation leads to attain relatively low MWD values and, if the process is not efficiently operated, there is always the possibility of obtaining a polymer network. In this work, it was found that process operation is greatly enhanced by the semi-batch addition of 1,4-butanediol and diamine, and the manipulation of the reactor temperature profile, allowing to obtain high molecular weights while avoiding the onset of the gelation point.     

Data‐driven adaptive nested robust optimization: General modeling framework and efficient computational algorithm for decision making under uncertainty

A novel data‐driven adaptive robust optimization framework that leverages big data in process industries is proposed. A Bayesian nonparametric model—the Dirichlet process mixture model—is adopted and combined with a variational inference algorithm to extract the information embedded within uncertainty data. Further a data‐driven approach for defining uncertainty set is proposed. This machine‐learning model is seamlessly integrated with adaptive robust optimization approach through a novel four‐level optimization framework. This framework explicitly accounts for the correlation, asymmetry and multimode of uncertainty data, so it generates less conservative solutions. Additionally, the proposed framework is robust not only to parameter variations, but also to anomalous measurements. Because the resulting multilevel optimization problem cannot be solved directly by any off‐the‐shelf solvers, an efficient column‐and‐constraint generation algorithm is proposed to address the computational challenge. Two industrial applications on batch process scheduling and on process network planning are presented to demonstrate the advantages of the proposed modeling framework and effectiveness of the solution algorithm. © 2017 American Institute of Chemical Engineers AIChE J, 63: 3790–3817, 2017     

Multiway discrete hidden Markov model‐based approach for dynamic batch process monitoring and fault classification

A new multiway discrete hidden Markov model (MDHMM)‐based approach is proposed in this article for fault detection and classification in complex batch or semibatch process with inherent dynamics and system uncertainty. The probabilistic inference along the state transitions in MDHMM can effectively extract the dynamic and stochastic patterns in the process operation. Furthermore, the used multiway analysis is able to transform the three‐dimensional (3‐D) data matrices into 2‐D measurement‐state data sets for hidden Markov model estimation and state path optimization. The proposed MDHMM approach is applied to fed‐batch penicillin fermentation process and compared to the conventional multiway principal component analysis (MPCA) and multiway dynamic principal component analysis (MDPCA) methods in three faulty scenarios. The monitoring results demonstrate that the MDHMM approach is superior to both the MPCA and MDPCA methods in terms of fault detection and false alarm rates. In addition, the supervised MDHMM approach is able to classify different types of process faults with high fidelity. © 2011 American Institute of Chemical Engineers AIChE J, 2012