化工过程的智能混合建模方法及应用

随着人工智能技术和配套数据系统的快速发展,化工过程建模技术达到了新的高度,将多个机理模型和数据驱动模型以合理的结构加以组合的智能混合建模方法,可以综合利用化工过程的第一性原理及过程数据,结合人工智能算法以串联、并联或者混联的形式解决化工过程中的模拟、监测、优化和预测等问题,建模目的明确,过程灵活,形成的混合模型有着更好的整体性能,是近年来过程建模技术的重要发展趋势。本文围绕近年来针对化工过程的智能混合建模工作进行了总结,包括应用的机器学习算法、混合结构设计、结构选择等关键问题,重点论述了混合模型在不同任务场景下的应用。指出混合建模的关键在于问题和模型结构的匹配,而提高机理子模型性能,获取高质量宽范围的数据,深化对过程机理的理解,形成更有效率的混合建模范式,这些都是现阶段提高混合建模性能的研究方向。     

Evolutionary polymorphic neural network in chemical process modeling

Evolutionary polymorphic neural network (EPNN) is a novel approach to modeling dynamic process systems. This approach has its basis in artificial neural networks and evolutionary computing. As demonstrated in the studied dynamic CSTR system, EPNN produces less error than a traditional recurrent neural network with a less number of neurons. Furthermore, EPNN performs networked symbolic regressions for input–output data, while it performs multiple step ahead prediction through adaptable feedback structures formed during evolution. In addition, the extracted symbolic formulae from EPNN can be used for further theoretical analysis and process optimization.     

Advances in modelling and analysis of chemical process systems

The paper reviews work on modelling in two constrasting areas: the performance of sieve-plates in distillation and complex fluid flows. It then attempts to draw some conclusions on advances in modelling techniques, in the light of the different experiences in these two fields.     

WBRPLSR方法及其在化工软测量中的应用

及时测定化工过程变量,对确保生产过程稳定、有效控制产品质量具有重要意义.基于实时样本数据,采用偏最小二乘方法,以分块递归的方式,为过程变量建立软测量模型.在分析时序数据特性的基础上,引入加权策略,并提出选定相关参数的方法步骤,推导构建了加权分块递归偏最小二乘回归方法（WBRPLSR）.将该法实际应用于某公司PTA装置溶剂脱水塔,为塔釡排出液H₂O含量建立软测量模型,效果良好.与已有方法相比,它提高了建模效率,改进了预测性能.     

文化差分进化算法及其在化工过程建模中的应用

提出了一种新的文化差分进化算法,该算法将差分进化算法作为文化算法的种群空间,在文化算法的信念空间和影响函数设计中提出了基于多种知识源的设计方法,通过多种知识指导差分进化的变异操作和交叉操作,使知识的表达和指导种群进化的能力得到加强。函数测试结果表明,基于知识机制的引入使得文化差分进化算法在寻优性能上比差分进化算法有了较大的提高,而对参数的敏感性却相对较小。将文化差分进化算法用于训练补偿模糊神经网络,建立乙烯精馏塔产品质量软测量模型。通过训练与泛化能力的比较结果表明,基于文化差分进化算法的补偿模糊神经网络软测量模型在建模精度和泛化性能上均优于常规补偿模糊神经网络、模糊神经网络以及采用遗传算法优化的模型,具有更好的应用前景。     

Development of mathematical models and modeling of chemical engineering systems under uncertainty

A method that consists of developing models of different types (deterministic, statistical, fuzzy, hybrid) for elements of chemical engineering systems with consideration for various types of available information (theoretical, statistical, fuzzy) and the combination of these models into a single suite was proposed for the creation of a suite of mathematical models for a chemical engineering system. The proposed approach was used to construct a suite of models for the principal apparatuses of the sulfur-recovery unit of Atyrau oil refinery.     

Review and classification of recent observers applied in chemical process systems

Observers are computational algorithms designed to estimate unmeasured state variables due to the lack of appropriate estimating devices or to replace high-priced sensors in a plant. It is always important to estimate those states prior to developing state feedback laws for control and to prevent process disruptions, process shutdowns and even process failures. The diversity of state estimation techniques resulting from intrinsic differences in chemical process systems makes it difficult to select the proper technique from a theoretical or practical point of view for design and implementation in specific applications. Hence, in this paper, we review the applications of recent observers to chemical process systems and classify them into six classes, which differentiate them with respect to their features and assists in the design of observers. Furthermore, we provide guidelines in designing and choosing the observers for particular applications, and we discuss the future directions for these observers.     

Koopman Lyapunov-based model predictive control of nonlinear chemical process systems

In this work, we propose the integration of Koopman operator methodology with Lyapunov-based model predictive control (LMPC) for stabilization of nonlinear systems. The Koopman operator enables global linear representations of nonlinear dynamical systems. The basic idea is to transform the nonlinear dynamics into a higher dimensional space using a set of observable functions whose evolution is governed by the linear but infinite dimensional Koopman operator. In practice, it is numerically approximated and therefore the tightness of these linear representations cannot be guaranteed which may lead to unstable closed-loop designs. To address this issue, we integrate the Koopman linear predictors in an LMPC framework which guarantees controller feasibility and closed-loop stability. Moreover, the proposed design results in a standard convex optimization problem which is computationally attractive compared to a nonconvex problem encountered when the original nonlinear model is used. We illustrate the application of this methodology on a chemical process example.     

Mathematical programming approaches to the synthesis of chemical process systems

This paper presents a review of advances that have taken place in the mathematical programming approach to process design and synthesis. A review is first presented on the algorithms that are available for solving MINLP problems, and its most recent variant, Generalized Disjunctive Programming models. The formulation of superstructures, models and solution strategies is also discussed for the effective solution of the corresponding optimization problems. The rest of the paper is devoted to reviewing recent mathematical programming models for the synthesis of reactor networks, distillation sequences, heat exchanger networks, mass exchanger networks, utility plants, and total flowsheets. As will be seen from this review, the progress that has been achieved in this area over the last decade is very significant.     

A superstructure representation,generation, and modeling framework for chemical process synthesis

We present a framework for the efficient representation, generation, and modeling of superstructures for process synthesis. First, we develop a new representation based on three basic elements: units, ports, and conditioning streams. Second, we present four rules based on “minimal” and “feasible” component sets for the generation of simple superstructures containing all feasible embedded processes. Third, in terms of modeling, we develop a modular approach, and formulate models for each basic element. We also present a canonical form of element models using input/output variables and constrained/free variables. The proposed methods provide a coherent framework for superstructure‐based process synthesis, allowing efficient model generation and modification. © 2016 American Institute of Chemical Engineers AIChE J, 62: 3199–3214, 2016     

Mathematical modeling and vector fields of dynamical systems in chemical technology of organic substances

Some aspects of mathematical modeling used in chemical technology have been considered. The most attention has been concentrated on the mathematical models in differential form, for which the investigation of common regularities of multidimensional vector fields of dynamical systems, based on which the driving force of the process is formed, represents a necessary step. The classification of vector fields used in particular in the case of consideration of chemical reactions and reactors, separation processes and combined reaction-mass transfer processes has been performed. The studies of vector fields, which are typical of the chemical technology process under discussion make it possible to meaningfully proceed to the numerical mathematical modeling of this process on a computer.     

Model-based reasoning in diagnostic expert systems for chemical process plants

A prototype expert system, called MODEX, for locating the cause(s) of a set of abnormalities in a chemical process id described. We discuss a methodology that aids the developement of expert systems which are process-independent, transparent in their reasoning, and capable of diagnosing a wide diversity of faults. The domain knowlege of the system is based on qualitative reasoning principles and captures physical interconnections between equipment units as well as causal relationships among process state variables. The inference strategy uses model-based reasoning for analyzing the plant behavior. Using a variant of the technique adopted from fault tree synthesis, an initially observed abnormal symptom is considered to be a top level event and a tree structure is constructed as the system searches for a basic event to which the fault can be traced. The diagnostic reasoning process is driven by a problem reduction strategy. The knowledge base is process-independent, thereby enhancing the generality of the expert system. Reasoning from first-principles with the aid of causal and fault models facilitates the diagnoses of novel or unanticipated faults. The system does not assume a single causal origin for all initially observed faults in the chemical process. Moreover, the system has the ability to locate multiple basic causes of a fault. The methodology also permits one to investigate the causal origins of multiple, unrelated, faults. The system provides explanations to user queries at various degrees of detail. Two test cases are discussed in detail.     

复杂化工过程系统的能值计算方法

针对复杂过程系统的生态化设计,研究系统内部及输出流股能值的计算策略和方法。基于系统能量或质量流程图,确定系统内外的能流、物流的分布和能量分配系数,通过建立复杂系统能值计算的循迹法,确定系统各种输入能值在系统内的分配,解决复杂系统内部联结各单元的流股能值准确计算问题。基于该方法,对玉米燃料乙醇生产过程系统进行了能值分析和评价。     

Environmentally benign chemical additives in the treatment and chemical cleaning of process water systems: Implications for green chemical technology

Supersaturated process waters high in silicates frequently result in deposition of colloidal silica or metal silicate salts. Silica cannot be inhibited by conventional phosphonate mineral scale inhibitors. Chemical cleaning poses hazards and requires operational shut-downs. This paper is focused on a dual approach for silica scale control, inhibition of colloidal silica formation and colloidal silica dissolution in water technology applications by use of designed chemical approaches. The additives used for silica inhibition were polyaminoamide dendrimers (PAMAM) and polyethyleneimine (PEI), in combination with carboxymethyl inulin (CMI) and polyacrylate (PAA) polymers. In principle, silica inhibition is a function of time and inhibitor dosage. Amine-terminated PAMAM-1 and 2 dendrimers as well as PEI combined with anionic polymers, such as CMI and PAA, seem to have a significant inhibitory effect on silica formation, most likely at its earlier stages where the reaction products are oligomeric silicates. CMI and PAA assist the inhibitory action of PAMAM-1 and 2 and PEI by alleviating formation of insoluble SiO₂-PAMAM precipitates. This most likely occurs by partial neutralization of the positive charge that exists in –NH⁺₃ surface groups. Increase of anionic polymer dosage above a certain threshold has a detrimental effect on the activity of the cationic inhibitors. In that case the polymer’s negative charge “overwhelms” the cationic charge of the inhibitor and poisons its inhibition ability. For silica dissolution, acetic, oxalic, citric acids, histidine and phenylalanine were used as potential replacements of ammonium bifluoride (NH₄F·HF). Silica dissolution is dependent in a rather unpredictable fashion on the structure of the dissolver, time and dosage. This paper continues our research efforts in the discovery, design and application of antiscalant additives that have mild environmental impact. These chemicals are also known as “green additives”.     

Fault monitoring based on mutual information feature engineering modeling in chemical process

A large amount of information is frequently encountered when characterizing the sample model in chemical process. A fault diagnosis method based on dynamic modeling of feature engineering is proposed to effectively remove the nonlinear correlation redundancy of chemical process in this paper. From the whole process point of view, the method makes use of the characteristic of mutual information to select the optimal variable subset. It extracts the correlation among variables in the whitening process without limiting to only linear correlations. Further, PCA (Principal Component Analysis) dimension reduction is used to extract feature subset before fault diagnosis. The application results of the TE (Tennessee Eastman) simulation process show that the dynamic modeling process of MIFE (Mutual Information Feature Engineering) can accurately extract the nonlinear correlation relationship among process variables and can effectively reduce the dimension of feature detection in process monitoring.     

Multiphase CFD-based models for chemical looping combustion process: Fuel reactor modeling

Chemical looping combustion (CLC) is a flameless two-step fuel combustion that produces a pure CO₂ stream, ready for compression and sequestration. The process is composed of two interconnected fluidized bed reactors. The air reactor which is a conventional circulating fluidized bed and the fuel reactor which is a bubbling fluidized bed. The basic principle is to avoid the direct contact of air and fuel during the combustion by introducing a highly-reactive metal particle, referred to as oxygen carrier, to transport oxygen from the air to the fuel. In the process, the products from combustion are kept separated from the rest of the flue gases namely nitrogen and excess oxygen. This process eliminates the energy intensive step to separate the CO₂ from nitrogen-rich flue gas that reduce the thermal efficiency.Fundamental knowledge of multiphase reactive fluid dynamic behavior of the gas-solid flow is essential for the optimization and operation of a chemical looping combustor.Our recent thorough literature review shows that multiphase CFD-based models have not been adapted to chemical looping combustion processes in the open literature. In this study, we have developed the reaction kinetics model of the fuel reactor and implemented the kinetic model into a multiphase hydrodynamic model, MFIX, developed earlier at the National Energy Technology Laboratory. Simulated fuel reactor flows revealed high weight fraction of unburned methane fuel in the flue gas along with CO₂ and H₂O. This behavior implies high fuel loss at the exit of the reactor and indicates the necessity to increase the residence time, say by decreasing the fuel flow rate, or to recirculate the unburned methane after condensing and removing CO₂.