融合过程先验知识的递归神经网络模型及其应用

大部分化工过程具有非线性特性,一般的线性建模方法难以有效应用。针对非线性化工过程动态建模,提出了一种基于过程先验知识的递归神经网络模型,充分发掘化工过程隐含的先验知识,并将这些先验知识以非线性约束的形式嵌入NARMAX结构的前馈神经网络中,同时基于增广拉格朗日乘子法约束处理机制,用PSO-IPOPT混合优化算法对过程先验知识递归神经网络权值进行优化。该过程先验知识递归神经网络模型对非线性化工过程动态建模,不仅有良好的建模精度和预测外推能力,而且能避免零增益的出现和增益反转,确保网络模型在实际应用中的安全性。文中以环管式丙烯聚合反应过程实际工业数据验证了所提网络模型的有效性。     

结合机理先验约束的Kriging模型参数优化方法研究

基于过程先验知识和Kriging模型,针对Kriging模型在求解参数过程中容易陷入局部最优的缺陷,提出一种利用进化优化算法求解模型参数的方法。在参数求解过程中,将过程变量对应的输出函数的单调性和凹凸性先验知识作为约束条件,在合理的样本数目下,得到更符合先验知识的Kriging模型,提高模型的预测精度。通过函数仿真实验和复合纳滤膜的溶胶粒径的预测估计验证了此建模方法的有效性。     

基于改进粒子群优化的连续系统闭环辨识新方法

将粒子群优化算法(PSO)的全局辨识能力以及无约束多维寻优法--Rosenbrock的局部搜索能力相结合提出一种全新的参数辨识方法.该方法不需要控制器的先验知识,对测试信号无任何要求.仿真实验表明,与PSO-SQP算法及两阶段闭环辨识法相比,该方法对连续系统在响应不充分的情况下有良好的辨识效果.     

神经网络建模方法在维生素C发酵过程中的应用

传统的标准“黑箱”型人工神经网络已较为广泛地应用于生化过程中的状态预估等多个方面，然而结合过程先验知识或部分机理模型的混合神经网络建模方法能给出更令人满意的结果.本文将其应用于2-酮基-l-古龙酸（2-KLG）发酵过程的状态估计，并将其结果与传统神经网络模型进行了比较，混合模型明显优于单一神经网络方法.     

基于热扩散核密度确定密度峰值法的历史工况识别

在工业生产过程中,生产决策的调整或生产状况的变化会导致生产过程多模态化,常用的数据聚类方法进行工况识别时存在参数选取困难或需要先验知识等限制。基于此,提出一种将人工智能领域的热扩散核密度确定密度峰的技术与高斯混合模型相结合的方法,可有效克服目前方法的缺点。该方法首先利用热扩散核密度确定密度峰的技术估算每个数据点的密度及其与局部密度较大点的距离,获取数据集的聚类中心并完成聚类;其次,利用高斯混合模型获取不同工况的特征参数：平均值、协方差和先验概率,从而对多工况历史过程进行准确的描述;最后,利用文献中仿真数据和Tennessee Eastman过程两个案例进行验证,并与K-均值法和F-J改进的高斯混合模型进行比较,证明了本文提出方法可更加方便、有效地对历史工况进行准确识别。     

基于混合逻辑动态法的污水处理除氮过程的优化控制

应用混合逻辑动态系统法（MLD）,对活性污泥1号模型（ASM1）进行了简化,利用有关活性污泥法的专家经验,建立了连续进水间歇曝气活性污泥法除氮动态模型.再采用预测控制的方法对该过程进行优化控制,通过仿真结果可知：将MLD法应用到活性污泥法建模和优化控制中,可以更加深入地挖掘相关的专家经验知识,并将这些专家经验知识和连续变量模型相结合,使模型更加精确,控制和优化的效果更好.为污水处理领域研究提供了一条新的途径.     

基于混合核函数PCR方法的工业过程软测量建模

针对核函数PCR方法中单个核函数的局限性，提出了混合核函数PCR方法并将其用于工业过程软测量建模，以提高模型的推广能力和非线性处理能力。混合核函数集中了多个核函数的特点，兼具局部和全局特性，并可以通过参数调节他们对混合核函数的作用，将过程的先验知识融入到核函数的确定，进而适合不同的应用。工业丙烯腈生产过程丙烯腈收率软测量建模的应用表明，基于混合核函数PCR的软测量模型要优于核函数PCR方法，满足了工业应用要求。     

神经网络建模方法在维生素C发酵过程中的应用

传统的标准“黑箱”型人工神经网络已较为广泛地应用于生化过程中的状态预估等多个方面,然而结合过程先验知识或部分机理模型的混合神经网络建模方法能给出更令人满意的结果.本文将其应用于2-酮基-l-古龙酸（2-KLG）发酵过程的状态估计,并将其结果与传统神经网络模型进行了比较,混合模型明显优于单一神经网络方法.     

对抗菌陶瓷材料性能及制备工艺的计算机定量评估

本文主要介绍了一种多目标决策分析方法-具有先验偏好的离散模型的消去与选择转换算法ELECTREI，并对该法用于抗菌陶瓷材料性能及制备工艺的评估作了详细讨论。最后，作者还对多目标决策分析方法的进一步应用前景作了展望。     

地质知识库与多点统计地质建模——以延10_(1~(2-1))、10_(1~(2-2))小层夹层建模为例

储层地质知识库的建立则是储层建模中一项十分重要的基础工作。储层地质知识库是通过充分利用已有的各种资料对研究区进行详细研究和统计分析,建立表征储层各种特征的地质知识。这些知识可以直接作为输入参数参与储层建模,或者为某些参数的确定、模拟方法的选择、实现的选取及结果的检验提供数据或地质依据。多点统计地质建模方法的最大优势就是可以将地质家的先验认识作为确定性的要素加入到建模过程中。地质知识库刚好可以提供这些先验的地质认识。本文以延1012-1、1012-2小层夹层建模为例,重点阐述如何将地质知识库的内容应用到地质建模中,提高地质知识库的实用性,探索出一条地质知识库与地质建模相结合的新途径。     

聚合物合成工艺学-教学浅谈

《高聚物合成工艺学》是工科材料化学专业重要的专业课程。根据聚合物合成工艺学的课程特点,结合课堂教学的经验,作者从学生知识体系的构建、知识的运用以及教学内容、教学方法阐述了教学体会,制定出了更加适合学生学习的教学方法和考核方式,以求在传授知识的同时,努力提高该课程的教学质量,激发学生的学习兴趣,培养学生综合运用知识的能力。     

基于先验知识的杂交方法的设计及其在化工问题中的应用

Three-layer feedforward networks have been widely used in modeling chemical engineering processes and prior-knowledge-based methods have been introduced to improve their performances.In this paper,we propose the methodology of designing better prior-knowledge-based hybrid methods by combining the existing ones. Then according to this methodology,two hybrid methods,interpolation-optimization (IO) method and interpolation penalty-function (IPF) method,are designed as examples.Finally,both methods are applied to modeling two cases in chemical engineering to investigate their effectiveness.Simulation results show that the performances of the hybrid methods are better than those of their parents.     

Maximum entropy estimation of the bubble size distribution in fluidized beds

This work presents a new methodology, based on the maximum entropy method, to obtain bubble characteristics in fluidized beds. The probability distributions (PDF) of bubble pierced length and velocity are obtained applying the maximum entropy principle to experimental measurements. In addition, the bubble diameter distribution has been inferred from experimental pierced length measurements. This method is applied to characterize bubbles in fluidized beds for the first time and the most general bubble geometry, a truncated spheroid, is considered. The distance between probes, s, which is the minimum pierced length that is possible to measure accurately using intrusive probes, has been introduced as a constraint in the derivation of the size distribution equation.The maximum entropy method is applied to experimental measurements of bubble characteristics carried out using optical and pressure probes in a three-dimensional fluidized bed of Geldart B particles. Results on bubble size obtained from pressure and optical probes are very similar, although optical probes provide more local information and can be used at any position in the bed. The maximum entropy principle has been found to be a simple method that offers many advantages over other methods applied before for size distribution modeling in fluidized beds.     

Automated synthesis of steady-state continuous processes using reinforcement learning

Automated flowsheet synthesis is an important field in computer-aided process engineering.The present work demonstrates how reinforcement learning can be used for automated flowsheet synthesis without any heuristics or prior knowledge of conceptual design.The environment consists of a steady-state flowsheet simulator that contains all physical knowledge.An agent is trained to take discrete actions and sequentially build up flowsheets that solve a given process problem.A novel method named SynGameZero is developed to ensure good exploration schemes in the complex problem.Therein,flowsheet synthesis is modelled as a game of two competing players.The agent plays this game against itself during training and consists of an artificial neural network and a tree search for forward planning.The method is applied successfully to a reaction-distillation process in a quaternary system.     

The characterization of nonionic surfactants by NMR

The methods currently reported in the literature for the characterization of nonionic surfactants are usually applied to one portion of the molecule and require a knowledge of the other portion for complete identification. This indirect approach leaves much to be desired. A simple, rapid, and more direct method of characterization is to measure the proton signal intensity in high resolution nuclear magnetic resonance (NMR) spectra. This method determines the hydrophile to hydrophobe ratio without requiring standard samples for calibration or a prior knowledge of the hydrophobe. In addition, this method will frequently give much valuable information about the identity of the hydrophobe, such as the average chain length, the degree of branching, and the type of aromatic substitution, if any. This method has been applied to the characterization of the common types of commercial polyethylene oxide condensates. The application of NMR to the analysis of formulated detergent products is also discussed. Presented at the AOCS meeting in New Orleans, La., 1962.     

Reconstruction of electrical capacitance tomography images based on fast linearized alternating direction method of multipliers for two-phase flow system

Electrical capacitance tomography (ECT) has been applied to two-phase flow measurement in recent years. Image reconstruction algorithms play an important role in the successful applications of ECT. To solve the il-posed and nonlinear inverse problem of ECT image reconstruction, a new ECT image reconstruction method based on fast lin-earized alternating direction method of multipliers (FLADMM) is proposed in this paper. On the basis of theoretical analysis of compressed sensing (CS), the data acquisition of ECT is regarded as a linear measurement process of permittivity distribution signal of pipe section. A new measurement matrix is designed and L1 regularization method is used to convert ECT inverse problem to a convex relaxation problem which contains prior knowledge. A new fast alternating direction method of multipliers which contained linearized idea is employed to minimize the objective function. Simulation data and experimental results indicate that compared with other methods, the quality and speed of reconstructed images are markedly improved. Also, the dynamic experimental results in-dicate that the proposed algorithm can fulfil the real-time requirement of ECT systems in the application.     

Toward an intelligent approach for predicting surface tension of binary mixtures containing ionic liquids

Knowledge of the surface tension of ionic liquids (ILs) and their related mixtures is of central importance and enables engineers to efficiently design new processes dealing with these fluids on an industrial scale. It’s obvious that experimental determination of surface tension of every conceivable IL and its mixture with other compounds would be a herculean task. Besides, experimental measurements are intrinsically laborious and expensive; therefore, accurate prediction of the property using a reliable technique would be overwhelmingly favorable. To do so, a modeling method based on artificial neural network (ANN) trained by Bayesian regulation back propagation training algorithm (trainbr) has been proposed to predict surface tension of the binary ILs mixtures. A total set of 748 data points of binary surface tension of IL systems within temperature range of 283.1-348.15 K was used to train and test the applied network. The obtained results indicated that the predictive values and experimental data are quite matching, representing reliability of the used ANN model for such purpose. Also, compared with other methods, such as SVM, GA-SVM, GA-LSSVM, CSA-LSSVM, GMDH-PNN and ANN trained with trainlm algorithm the proposed model was better in terms of accuracy.     

关键核网络及其在发酵过程在线建模中的应用

发酵过程通常采用流加补料操作,无稳态工作点、非线性强,且重要生物量往往无法在线测量。本文提出了一种适用于非线性多输入多输出的发酵过程在线建模方法：关键核网络（key kernel network,KKN）。结合过程的先验知识给出控制模型关键节点加入的准则,使其能自适应调整模型的复杂度,以提高建模的精度和速度,并给出了关键节点增加时KKN模型的在线递推形式。将KKN应用于青霉素发酵过程的在线建模,研究表明,KKN能同时快速、准确地预报菌体和产物浓度,且随着批次的增加,过程信息不断得到积累,模型精度逐渐提高。     

冷却水中藻类去除方法的研究进展

藻类污染作为水体污染的一部分,已引起人们的广泛关注。作者对物理法、化学法和生物法去除冷却水中的藻类的国内外现状进行了概述。综述了光磁协同、电化学等物理除藻方法,铜电解、臭氧氧化、化学药剂复配等化学除藻方法,以及利用水网藻和改性海带的生物除藻方法。并对各种方法的除藻机理及应用效果进行了简述。在总结国内外研究的基础上,对未来发展前景进行了评价和展望。