模糊自适应PID算法在核磁共振谱仪样品旋转控制系统中的应用

在核磁共振(NMR)实验中,样品旋转能够有效消除 XY 方向磁场的部分不均匀性,从而提高信号分辨率.在商用NMR谱仪中,一般采用PID算法进行待测样品的旋转控制.由于被控对象具有一定的非线性,存在着调节时间较长,稳定后存在误差等缺点.针对该现象,在Varian 500 MHz波谱仪气路系统的基础上,设计一套新的控制电路...     

Effects of multiple-stage membrane process designs on the achievable performance of automatic control

There is limited information from literature on the dynamic operability of membrane processes with multiple stages or loops. Such information is useful for assessing the performance achievable by an automatic controller proposed for a process design before the actual controller is implemented. Based on dynamic modeling of an industrial whey ultrafiltration process with an increasing number of stages up to a maximum of 12, the dynamic responses of the flowrate and concentration of the retentate were obtained. Features of the dynamic responses were used to determine the performance, in terms of quality and speed, that can be achieved by automatic controllers. In particular, limitations on the performance are indicated by features of dynamic responses such as effective time delay and inverse responses. Changes in effective time delay and inverse responses with the number of stages in the whey ultrafiltration process demonstrate a trade-off between process performance and control performance. This trade-off should be considered during process and controller design to maximize the economic return from the production of whey protein concentrates.     

Fuzzy self-tuning PID control of the operation temperatures in a two-staged membrane separation process

A two-staged membrane separation process for hydrogen recovery from refinery gases is introduced. The principle of the gas membrane separation process and the influence of the operation temperatures are analyzed. As the conventional PID controller is difficult to make the operation temperatures steady, a fuzzy self-tuning PID control algorithm is proposed. The application shows that the algorithm is effective, the operation temperatures of both stages can be controlled steadily, and the operation flexibility and adaptability of the hydrogen recovery unit are enhanced with safety. This study lays a foundation to optimize the control of the membrane separation process and thus ensure the membrane performance.     

A Simple Substrate Feeding Strategy using a pH Control Trigger in Fed-Batch Fermentation

A simple automated glucose feeding strategy based on pH control was developed to produce high-cell-density fed-batch fermentation. In this strategy, the pH control scheme utilized an acidified concentrated glucose solution to lower the pH. The frequency of glucose addition to the fermentor is determined by the culture’s growth kinetics. To demonstrate the effectiveness of the coupled pH and glucose control strategy in biomass and/or secondary metabolite production, several fed-batch fermentations of indigenous Escherichia coli and recombinant E. coli were carried out. Both strains produced biomass with optical density of greater than 40 at 600 nm. We also tested the glucose control strategy using two types of pH controller: a less sophisticated portable pH controller and a more sophisticated online proportional-integral-derivative (PID) controller. Our control strategy was successfully applied with both controllers, although better control was observed using the PID controller. We have successfully demonstrated that a glucose feeding strategy based on a simple pH control scheme to indirectly control the glucose concentration can be easily achieved and adapted to conventional bioreactors in the absence of online glucose measurement and control.     

用于药品质量快速检测的近红外光谱模糊神经元分类方法

针对非线性且分类界线模糊的药品质量类别快速测定难题, 将近红外光谱分析与模糊神经网络相结合, 经研究提出近红外光谱模糊神经网络分类方法, 用于计算辨析中药等化学组成复杂药品的近红外光谱模式类别, 从而快速评定这类药品的质量. 以参麦注射液为典型分析对象, 以鉴别其生产厂家这一模式分类问题为例, 考核本文方法, 结果表明, 其分类准确率达到94.2%, 明显优于经典的BP神经网络分类方法(84.6%), 可望用于中药产品质量类别的快速检测与评价.     

一类基于模糊神经元的复杂非线性化学模式识别方法

针对模式类别边界曲折而模糊的复杂化学模式分类问题,提出一种化学模式模糊分类方法,并给出其模糊神经元分类器设计和网络训练算法,使模糊神经元分类器具有学习功能.以一个应用实例检验了该方法的实效.     

Experimental implementation of automatic ‘cycle to cycle’ control to a nonlinear chiral simulated moving bed separation

In order to better exploit the economic potential of the simulated moving bed chromatography a ‘cycle to cycle’ controller which only requires the information about the linear adsorption behavior and the overall average porosity of the columns has been proposed. Recently, an automated on-line HPLC monitoring system which determines the concentrations in the two product streams averaged over one cycle, and returns them as feedback information to the controller was implemented. The new system allows for an accurate determination of the average concentration of the product streams even if the plant is operated at high concentrations. This paper presents the experimental implementation of the ‘cycle to cycle’ control concept to the separation of guaifenesin enantiomers under nonlinear chromatographic conditions, i.e. at high feed concentrations. Different case studies have been carried out to challenge the controller under realistic operation conditions, e.g. introducing pump disturbances and changing the feed concentration during the operation. The experimental results clearly demonstrate that the controller can indeed deliver the specified purities and improve the process performance.     

Design and Evaluation of an Optimal Controller for Simultaneous Saccharification and Fermentation Process

Ethanol from corn is produced using dry grind corn process in which simultaneous saccharification and fermentation (SSF) is one of the most critical unit operations. In this work an optimal controller based on a previously validated SSF model was developed by formulating the SSF process as a Bolza problem and using gradient descent methods. Validation experiments were performed to evaluate the performance of optimal controller under different process disturbances that are likely to occur in practice. Use of optimal control algorithm for the SSF process resulted in lower peak glucose concentration, similar ethanol yields (13.38±0.36% v/v and 13.50±0.15% v/v for optimally controlled and baseline experiments, respectively). Optimal controller improved final ethanol concentrations as compared to process without optimal controller under conditions of temperature (13.35±1.28 and 12.52±1.19% v/v for optimal and no optimal control, respectively) and pH disturbances (12.65±0.74 and 11.86±0.49% v/v for optimal and no optimal control, respectively). Cost savings due to lower enzyme usage and reduced cooling requirement were estimated to be up to $1 million for a 151 million L/yr (40 million gal/yr) dry grind plant.     

Plasma Spray Process Operating Parameters Optimization Based on Artificial Intelligence

During plasma spray process, many intrinsic operating parameters allow tailoring in-flight particle characteristics (temperature and velocity) by controlling the plasma jet properties, thus affecting the final coating characteristics. Among them, plasma flow mass enthalpy, flow thermal conductivity, momentum density, etc. result from the selection of extrinsic operating parameters such as the plasma torch nozzle geometry, the composition and flow rate of plasma forming gases, the arc current intensity, beside the coupled relationships between those operating parameters make difficult in a full prediction of their effects on coating properties. Moreover, temporal fluctuations (anode wear for example) require “real time” corrections to maintain particle characteristic to targeted values. An expert system is built to optimize and control some of the main extrinsic operating parameters. This expert system includes two parts: (1) an artificial neural network (ANN) which predicts an extrinsic operating window and (2) a fuzzy logic controller (FLC) to control it. The paper details the general architecture of the system, discusses its limits and the typical characteristic times. The result shows that ANN can predict the characteristics of particles in-flight from coating porosity within maximal error 3 and 2 % in temperature and velocity respectively. And ANN also can predict the operating parameters from in-flight particle characteristics with maximal error 2.34, 4.80 and 8.66 % in current intensity, argon flow rate, and hydrogen flow rate respectively.     

自适应模糊偏最小二乘方法在药物构效关系建模中的应用

作为一种局部逼近方法,自适应神经模糊推理系统(ANFIS)适于为药物定量构效关系(QSAR)建模。描述药物分子结构的参数较多,常存在耦合关系,会增加建模难度,并影响模型的预报性能。为此,将ANFIS和偏最小二乘(PLS)相结合,先由PLS从样本数据中提取成分,再由ANFIS实现每对成分间的非线性映射,并基于输出误差进一步修正所提取的成分,使之对因变量具有最优的解释能力,由此构建为EB-AFPLS方法。该法已成功地应用于HIV-1蛋白酶抑制剂的QSAR建模,效果良好,显示出很强的学习能力,所建模型的预报性能也优于其它方法。     

运用模糊神经网络表达和预测链烷烃pVT性质

采用一种基于遗传算法的新型模糊神经网络方法研究链烷烃类化合物的pVT性质。该方法综合神经网络、遗传算法与模糊系统三种柔性智能计算技术的优点,具有良好的学习能力,不易陷入局部最小区域,学习速度较快,网络知识以模糊语言变量的形式加以表达,易于理解。用分子连接性指数对24种链烷烃化合物结构和pVT数据进行学习,进而预测另外14种未知化合物的pVT性质,较好地揭示出化合物分子结构与pVT性质之间的关系,并给出了良好的关联与预测结果。     

Experimental implementation of automatic 'cycle to cycle' control of a chiral simulated moving bed separation

In the absence of a suitable controller, currently simulated moving beds (SMBs) are operated suboptimally to cope with system uncertainties and to guarantee robustness of operation. Recently, we have developed a 'cycle to cycle' optimizing controller that not only makes use of minimal system information, i.e. only the Henry constants and average bed voidage, but also optimizes the process performance and taps the full economic potential of the SMB technology. The experimental implementation of the 'cycle to cycle' optimizing controller had been carried out for achiral separation. For chiral separation however, application of any online controller has not been possible because an appropriate online monitoring system has not been available. This work reports and discusses the first experimental implementation of the 'cycle to cycle' optimizing control for chiral separations. A mixture of guaifenesin enantiomers is separated on Chiralcel OD columns with ethanol as mobile phase in a eight-column four sections laboratory SMB unit. The results show that the controller, although using minimal information about the retention of the two enantiomers, is able to meet product and process specifications, can optimize the process performance, and is capable of rejecting disturbances that may occur during the operation of the SMB plant.     

Modeling of the relationship between electroosmotic flow and separation parameters in capillary zone electrophoresis using artificial neural networks and experimental design

The prediction of migration time of electroosmotic flow (EOF) marker was achieved by applying artificial neural networks (ANN) model based on principal component analysis (PCA) and standard normal distribution simulation to the input variables. The voltage of performance, the temperature in the capillary, the pH and the ionic strength of background electrolytes (BGE) were applied as the input variables to ANN. The range of the performance voltage studied was from 15 to 27 kV, and that of the temperature in the capillary was from 20 to 30 °C. For the pH values studied, the range was from 5.15 to 8.04. The range of the ionic strength investigated in this paper was from 0.040 to 0.097. The prediction abilities of ANN with different pre-processing procedure to the input variables were compared. Under the same performance conditions, the average prediction error of the migration time of the EOF marker was 5.46% with RSD = 1.76% according to 10 parallel runs of the optimized ANN structure by the proposed approach, and that of the 10 parallel predictions of the optimal ANN structure for the different performance conditions was 12.95% with RSD = 2.29% according to the proposed approach. The study showed that the proposed method could give better predicted results than other approaches discussed.     

Analyses of the temperature and pH effects on the complexation of magnesium and calcium in human blood plasma: an approach using artificial neural networks

The temperature and pH effects on the equilibrium of a blood plasma model have been studied on the basis of artificial neural networks. The proposed blood plasma was modeled considering two important metals, calcium and magnesium, and six ligands, namely, alanate, carbonate, citrate, glycinate, histidinate and succinate. A large data set has been used to simulate different concentrations of magnesium and calcium as a function of temperature and pH and these data were used for training the neural network. The proposed model allowed different types of analyses, such as the effects of pH on calcium and magnesium concentrations, the competition between calcium and magnesium for ligands and the effects of temperature on calcium and magnesium concentrations. The model developed was also used to predict how the variation of calcium concentration can affect magnesium concentrations. A comparison of neural network predictions against experimental data produced errors of about 3%. Moreover, in agreement with experimental measurements (Wang et al. in Arch. Pathol. 126:947–950, 2002; Heining et al. in Scand. J. Clin. Lab. Invest. 43:709–714, 1983), the artificial neural network predicted that calcium and magnesium concentrations decrease when pH increases. Similarly, the magnesium concentrations are less sensitive than calcium concentrations to pH changes. It is also found that both calcium and magnesium concentrations decrease when the temperature increases. Finally, the theoretical model also predicted that an increase of calcium concentrations will lead to an increase of magnesium concentration almost at the same rate. These results suggest that artificial neural networks can be efficiently applied as a complementary tool for studying metal ion complexation, with especial attention to the blood plasma analysis. Figure Artificial neural networks for predicting the behavior of calcium and magnesium as a function of pH and temperature in human blood plasma     

Prediction of coefficients of the Langmuir adsorption isotherm using various artificial intelligence (AI) techniques

Adsorption is a process that utilizes porous solid materials to separate some solutes from gas or liquid mixtures. The extent of this separation is often determined using the adsorption isotherms, i.e., semi-empirical correlation for relating the amount of adsorbed substances by the solid medium to its associated concentration in fluid phase at constant temperature. Prior to employing an adsorption isotherm, its coefficients should be adjusted using experimental data of a considered adsorption system. In this study, the coefficients of Langmuir model have been predicted using various types of artificial neural networks (ANNs), support vector machines, and adaptive neuro fuzzy interface systems, and coupled scheme of ANN-genetic algorithm. The employed ANN types are multi-layer perceptron neural network (MLPNN), radial basis function neural network, cascade feedforward neural network, and generalized neural network. The considered coefficients tried to be modeled as functions of temperature, pH, adsorbent density, and adsorbate molecular weight. Predictive accuracies of the AI techniques have been compared utilizing different statistical indices such as correlation coefficient (R²), mean square error, and absolute average relative deviation (AARD%). The results indicated that MLPNN was the most accurate model for predicting the coefficients of Langmuir isotherm, due to its AARDs of 24.64 and 22.40% for the first and second coefficients, respectively.     

辨识药物定量构效关系的模糊神经网络方法研究

提出一种基于遗传算法的新型模糊神经网络方法,用于计算Benzodiazepines(BZs)类药物的定量构效关系.这类模糊神经网络综合了神经网络、遗传算法与模糊逻辑的各自优势,具有优良的定量构效关系辨识能力,其学习速度较快,不易陷入局部最小区域;网络知识以模糊语言变量的形式加以表达,不仅易于理解,而且能有效地利用已有的专家经验.一旦通过学习获得规律后,不仅能很好地预测化合物的活性,还能对后续的药物分子设计提供有益的理论指导.     

A novel adaptive neuro-fuzzy inference system model to predict the intrinsic mechanical properties of various cellulosic fibers for better green composites

In this work a novel adaptive neuro-fuzzy inference system model has been developed for the prediction of the intrinsic mechanical properties of various cellulosic natural fibers to enhance their selection for better green composite materials. The model combined modeling function of the fuzzy inference system with the learning capability of the artificial neural network. The developed model was built up based on experimental mechanical properties of various cellulosic fiber types commonly used for natural fiber reinforced composites, and the rules have been generated directly from the experimental data. The developed model was capable of predicting all of Young's modulus, ultimate tensile strength, and elongation at break properties from only two intrinsic properties of fibers namely; cellulose and moisture content. The adaptive neural fuzzy inference system (ANFIS) structure included five layers to realize the establishment and calculation of each model. The system architecture included the fuzzy input layer, product layer, normalized layer, de-fuzzy layer and total output layer. Results have been revealed that the model’s predictions were highly in agreement with other experimentally gained properties when compared with experimental results for verifying the approach. The accuracy of the developed model would enhance predicting other cellulosic fiber properties to develop better natural fiber composites in the near future.

     

Quantitative fuzzy neural network for analytical determination

A quantitative fuzzy neural network (Q-FNN) for pattern recognition in analytical determination is reported in this paper. The fuzzy neural network (FNN) combines a fuzzy logic system with an artificial neural network (ANN) so that it has both advantages of a high training speed and strong anti-interference. Importantly, the analytical concept of relative error (RE) in quantitative determination has been integrated into FNN so that the Q-FNN provides a very good quantitative capability in chemical analysis, and prevents the system from an over-fitting problem. The logarithm curve with noise in terms of analytical response versus concentration is calibrated by trained FNN and a close approximation to the ideal one without noise is obtained. The Q-FNN has been applied to the concentration determination of freon in the presence of interference gases. The prediction error for a test set in quantification is less than 10% while no qualitative mistake is observed, implying that the quantitative FNN has sustained the feature of pattern recognition. The results indicate that the Q-FNN has obvious advantages not only in converging speed, but also in the quantitative accuracy over the ANN.     

A neuro-fuzzy system for X-ray spectra interpretation

A decision scheme for the interpretation of spectra from wavelength dispersive X-ray fluorescence spectrometry is described that encompasses elements from three areas of artificial intelligence: fuzzy logic, rule based expert systems and neural net technology.After transforming the recorded spectra to line spectra by appropriate background correction a reasoning scheme is applied that takes into account not only the observed spectra, but also the recording conditions and prior spectroscopic information regarding the relative emission probabilities and the usefulness of the different lines for the purpose of element identification. The latter is done on the basis of a previously described scheme to compute conditional a posteriori Bayes probabilities for a mean matrix. These different pieces of information are then assembled into a battery of fuzzy rules. The importance of the rules as well as the importance of the X-ray lines is determined in a training process, similar to the one in a feedforward back-propagation network.To further stabilize the results this network is pruned in a second training cycle. This, however, had little effect on the quality of interpretation.The advantages of this approach to the interpretation of X-ray spectra over older ones are numerous: the system adapts itself to better interpret spectra that are of greater importance to a laboratory as these are better represented in the training set; the fuzzy logic is capable of working with incomplete and uncertain knowledge, and the neural network results based on these fuzzy rules is readily interpretable by the X-ray spectroscopist as every rule can be expressed also in natural language as in any classical rule based system.On leave from Silesian University, Katowice, Poland     

Ru(III)-EDTA catalyzed oxidation of ascorbic acid by hydrogen peroxide in aqueous solution

The kinetics of oxidation of ascorbic acid to dehydroascorbic acid by hydrogen peroxide catalyzed by ethylenediaminetetraacetatoruthenate(III) has been studied over the pH range 1.50 – 2.50, at 30°C and μ = 0.1 M KNO₃. The reaction has a first-order dependence on ascorbic acid and Ru(III)-EDTA concentrations, an inverse first-order dependence on hydrogen ion concentration, and is independent of hydrogen peroxide concentration in the pH range studied. A mechanism has been proposed in which ascorbate anion forms a kinetic intermediate with the catalyst in a pre-equilibrium step. Ruthenium(III) is reduced to ruthenium(II) in a rate-determining step and is reoxidized with hydrogen peroxide back to the Ru(III) complex in a fast step.