基于支持向量机的发酵过程生物量在线估计

Biomass is a key factor in fermentation process, directly influencing the performance of the fermentation system as well as the quality and yield of the targeted product. Therefore, the on-line estimation of biomass is indispensable. The soft-sensor based on support vector machine （SVM） for an on-line biomass estimation was analyzed in detail, and the improved SVM called the weighted least squares support vector machine was presented to follow the dynamic feature of fermentation process. The model based on the modified SVM was developed and demonstrated using simulation experiments.     

A Selective Moving Window Partial Least Squares Method and Its Application in Process Modeling

A selective moving window partial least squares (SMW-PLS) soft sensor was proposed in this paper and applied to a hydro-isomerization process for on-line estimation of para-xylene (PX) content. Aiming at the high frequency of model updating in previous recursive PLSmethods, a selective updating strategywas developed. Themodel adaptation is activated once the prediction error is larger than a preset threshold, or themodel is kept unchanged. As a result, the frequency of model updating is reduced greatly,while the change of prediction accuracy is minor. The performance of the proposedmodel is better as compared with that of other PLS-based model. The compromise between prediction accuracy and real-time performance can be obtained by regulating the threshold. The guidelines to determine the model parameters are illustrated. In summary, the proposed SMW-PLS method can deal with the slow time-varying processes effectively.     

A study on the prediction of ozone formation in air pollution

Two prediction schemes-time series analysis and parameter estimation method-were investigated to predict the formation of ozone in Seoul, Korea. Moving average method and double exponential smoothing method are applied to the time-series analysis. Three typical methods, such as extended least squares (ELS), recursive maximum likelihood (RML) and generalized least squares (GLS), were used to predict ozone formation in a real time parameter estimation. Autoregressive moving average model with external input (ARMAX) is used as the model of the parameter estimation. To test the performance of the ozone formation prediction schemes proposed in the present work, the prediction results of ozone formation were compared to the real data. From the comparison it can be seen that the prediction scheme based on the parameter estimation method gives a reasonable accuracy with limited prediction horizon.     

A robust pole-placement self-tuning controller for chemical process control: Theoretical studies and simulation

A single variable pole-placement self-tuning controller (PPSTC) is used to simulate examples typical of chemical processes; i.e., open-loop stable, unstable, and unstable non-minimum phase systems with unknown varying process dead time. The PPSTC is shown to be effective in each case. Set-point tracking and rejection of randomly occurring deterministic disturbances for all three types of processes are achieved. Simultaneous estimation of process parameters and process time delay is realized by using a recursive extended least squares method.     

A soft sensor for flotation plants

A soft-sensor was developed to estimate the solids flow rates in the streams of a flotation circuit. The algorithm uses metal assays and mass flow rate measurements to perform the estimation. The method performs a recursive least squares estimation of the mass flow rates based on the assumption of mass conservation over a specified time window. The soft-sensor is programmed into a Vax computer system interfaced with a DCS and is currently used for process control and instrument tuning.     

Recursive least squares based estimation schemes for self-tuning control

Recursive Least Squares (RLS) is the most popular parametric identification method used for on-line process model estimation and self-tuning control. The basic least squares scheme is outlined in this paper and its lack of ability to track changing process parameters is illustrated and explained. Several variants of the basic algorithm which have appeared elsewhere in the literature are discussed. Some of these algorithms contain different modifications to the basic scheme which are intended to prevent this loss of alertness to changing process parameters. Other variations of the least squares algorithm are presented which attempt to deal with parameter estimation in the presence of disturbances and unmodelled process dynamics.     

利用状态空间模型和不敏卡尔曼滤波的分批补料发酵过程在线估计(英文)

On-line estimation of unmeasurable biological variables is important in fermentation processes, directly influencing the optimal control performance of the fermentation system as well as the quality and yield of the targeted product. In this study, a novel strategy for state estimation of fed-batch fermentation process is proposed. By combining a simple and reliable mechanistic dynamic model with the sample-based regressive measurement model, a state space model is developed. An improved algorithm, swarm energy conservation particle swarm optimization (SECPSO), is presented for the parameter identification in the mechanistic model, and the support vector machines (SVM) method is adopted to establish the nonlinear measurement model. The unscented Kalman filter (UKF) is designed for the state space model to reduce the disturbances of the noises in the fermentation process. The proposed on-line estimation method is demonstrated by the simulation experiments of a penicillin fed-batch fermentation process.     

氧化铝蒸发过程的混合智能优化设定控制

氧化铝蒸发过程的关键工艺指标碱液浓度不能在线检测且与控制回路输出之间的动态特性难以用精确的数学模型描述,采用已有的优化控制方法不能实现上述运行层的优化。针对上述问题提出了由基于案例推理的预设定模型,基于专家规则的前馈、反馈补偿模型以及基于块式偏最小二乘的软测量模型组成的混合智能优化设定控制方法。采用某氧化铝厂蒸发过程的实际数据进行仿真实验,实验研究表明该方法可以有效地将碱液浓度控制在工艺要求的区间内。     

一种用于时变系统的复位变参数递推最小二乘算法研究

Based on the idea of the set-membership identification,a modified recursive least squares algorithm with variable gain, variable forgetting factor and resetting is presented.The concept of the error tolerance level is proposed.The selection criteria of the error tolerance level are also given according to the min-max principle.The algorithm is particularly suitable for tracing time-varying systems and is similar in computational complexity to the standard recursive least squares algorithm.The superior performance of the algorithm is verified via simulation studies on a dynamic fermentation process.     

Adaptive multirate estimation and control of nutrient levels in a fed-batch fermentation using off-line and on-line measurements

A multirate adaptive estimation algorithm developed earlier (Gudi et al., 1995) is extended to perform estimation of nutrient levels using frequent on-line measurements of the carbon dioxide evolution rate (CER) and off-line, infrequent and delayed measurements of the biomass and substrate concentrations. It has been shown that the algorithm can be designed to track changing substrate yield coefficients as well. The estimation algorithm has been verified using simulations and industrial data from a fed-batch fermentation involving a Streptomyces specie. It has been coupled with a nonlinear control law designed to track prespecified optimal nutrient trajectories. The resulting closed loop control scheme is evaluated using simulation runs.     

Prediction and quality control of the melt index during production of high-density polyethylene

In polyolefin processes the melt index (MI) is the most important control variable indicating product quality. Because of the difficulty in the on-line measurement of MI, a lot of MI estimation and correlation methods have been proposed. In this work a new dynamic MI estimation scheme is developed based on system identification techniques. The empirical MI estimation equation proposed in the present study is derived from the 1 ^st -order dynamic models. Effectiveness of the present estimation scheme was illustrated by numerical simulations based on plant operation data including grade change operations in high density polyethylene (HDPE) processes. From the comparisons with other estimation methods it was found that the proposed estimation scheme showed better performance in MI predictions. The virtual sensor model developed based on the estimation scheme was combined with the virtual on-line analyzer (VOA) to give a quality control system to be implemented in the actual HDPE plant. From the application of the present control system, significant reduction of transition time and the amount of off-spec during grade changes was achieved     

A strategy to regulate continuous fermentation processes with unknown reaction rates

Identification and control of continuous fermentation processes are dif-ficult tasks due to the complexity and high coupling of dynamic behaviour of this kind of system. In this work is implemented an on-line estimation technique of the main uncertainties of a fermentation processes (e.g. specific growth rate, biomass concentration and yield coefficient) based on a mass balance, to generate a linearising feedback control law that provides a robust stabilisation against uncertainties. By numerical simulations the performance of the closed-loop system and the controller design procedure is illustrated.     

RECURSIVE GENERALIZED M ESTIMATES FOR AUTOREGRESSIVE MOVING-AVERAGE MODELS

Abstract. Outliers in time series seriously affect conventional parameter estimates. In this paper a robust recursive estimation procedure for the parameters of auto-regressve moving-average models with additive outliers is proposed. Using 'cleaned' residuals from an initial robust fit of an autoregression of high order as input, bounded influence regression is applied recursively. The proposal follows certain ideas of Hannan and Rissanen, who suggested a three-stage procedure for order and parameter estimation in a conventional setting.
A Monte Carlo study is performed to investigate the robustness properties of the proposed class of estimates and to compare them with various other suggestions, including least squares, M estimates, residual autocovariance and truncated residual autocovariance estimates. The results show that the recursive generalized M estimates compare favourably with them. Finally, possible modifications to master even vigourous situations are suggested.     

Three-way data analysis with time lagged window for on-line batch process monitoring

In this paper, on-line batch process monitoring is developed on the basis of the three-way data structure and the time-lagged window of process dynamic behavior. Two methods, DPARAFAC (dynamic parallel factor analysis) and DTri-PLS (dynamic trilinear partial least squares), are used here depending on the process variables only or on the process variables and quality indices, respectively. Although multivariate analysis using such PARAFAC (parallel factor analysis) and Tri-PLS (trilinear partial least squares) models has been reported elsewhere, they are not suited for practicing on-line batch monitoring owing to the constraints of their data structures. A simple modification of the data structure provides a framework wherein the moving window based model can be incorporated in the existing three-way data structure to enhance the detectability of the on-line batch monitoring. By a sequence of time window of each batch, the proposed methodology is geared toward giving meaningful results that can be easily connected to the current measurements without the extra computation for the estimation of unmeasured process variables. The proposed method is supported by using two sets of benchmark fault detection problems. Comparisons with the existing two-way and three-way multiway statistical process control methods are also included.     

ESTIMATION FOR NON-LINEAR TIME SERIES MODELS USING ESTIMATING EQUATIONS

Godambe's (1985) theorem on optimal estimating equations for stochastic processes is applied to non-linear time series estimation problems. Examples are considered from the usual classes of non-linear time series models. A recursive estimation procedure based on optimal estimating equations is provided. It is also shown that pre-filtered estimates can be used to obtain the optimal estimate from a non-linear state-space model.     

Recursive subspace identification with prior information using the constrained least squares approach

It is essential to develop high quality models for process control and other applications. The incorporation of prior information in subspace identification has been investigated to obtain improved model quality. One of the recent developments incorporates the prior information using the constrained least squares (CLS). In many online applications, the amount of process data for model identification grows with time, and it is therefore necessary to develop a recursive algorithm for online identification of process models and to address the time-varying characteristics of the systems. In this paper, a recursive subspace identification algorithm incorporating prior information is developed using the constrained recursive least squares (CRLS). It is shown via a simulation example that the state space model identified using the proposed algorithm possesses improved accuracy.     

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。     

ON-LINE OPTIMAL FEED OF SUBSTRATE DURING FED-BATCH CULTURE OF CELL MASS

In the cultivation of cell mass, a mathematical model for the specific growth rate of the cells can be used to design a strategy for the optimal feed of substrate to the culture. In this paper, a model of specific growth rate valid at both low and high cell concentration is developed for Candida utilis and Candida brassicae. An on-line control algorithm is then proposed to determine the optimal substrate concentration for the production of cell mass in fed-batch culture. An on-line estimation scheme using the sensitivity method is also included for cases in which the cell mass can not be measured. Simulation results demonstrate that the proposed method can minimize the production time of fed-batch culture.     

Nonlinear state estimation for fermentation process using cubature Kalman filter to incorporate delayed measurements☆

State estimation of biological process variables directly influences the performance of on-line monitoring and op-timal control for fermentation process. A novel nonlinear state estimation method for fermentation process is proposed using cubature Kalman filter (CKF) to incorporate delayed measurements. The square-root version of CKF (SCKF) algorithm is given and the system with delayed measurements is described. On this basis, the sample-state augmentation method for the SCKF algorithm is provided and the implementation of the proposed algorithm is constructed. Then a nonlinear state space model for fermentation process is established and the SCKF algorithm incorporating delayed measurements based on fermentation process model is presented to implement the nonlinear state estimation. Finally, the proposed nonlinear state estimation methodology is applied to the state estimation for penicillin and industrial yeast fermentation processes. The simulation results show that the on-line state estimation for fermentation process can be achieved by the proposed method with higher esti-mation accuracy and better stability.     

Development of real-time state estimators for reaction–separation processes: A continuous flash fermentation as a study case

The development of reliable on-line state estimators applicable to reaction–separation processes is addressed in this work. Artificial Neural Network-based software sensors (ANN-SS) are proposed to allow on-line measurement of key variables, with an estimation algorithm that uses secondary variables as inputs. A continuous laboratory-scale flash fermentation for bioethanol production is considered as a case study. The process consists of three interconnected units: fermentor, filter (tangential microfiltration for cell recycling) and vacuum flash vessel (for the continuous separation of ethanol from the broth). The concentrations of ethanol in the fermentor and of ethanol condensed from the flash are successfully monitored on-line using ANN-SS. The proposed model contributes to improve the understanding of the complex relationships between process variables in the reaction and separation units, which is of major importance to allow the operation of the ethanol production process near its optimum performance.