Neural‐Networks‐Based Feedback Linearization versus Model Predictive Control of Continuous Alcoholic Fermentation Process

In this work advanced nonlinear neural networks based control system design algorithms are adopted to control a mechanistic model for an ethanol fermentation process. The process model equations for such systems are highly nonlinear. A neural network strategy has been implemented in this work for capturing the dynamics of the mechanistic model for the fermentation process. The neural network achieved has been validated against the mechanistic model. Two neural network based nonlinear control strategies have also been adopted using the model identified. The performance of the feedback linearization technique was compared to neural network model predictive control in terms of stability and set point tracking capabilities. Under servo conditions, the feedback linearization algorithm gave comparable tracking and stability. The feedback linearization controller achieved the control target faster than the model predictive one but with vigorous and sudden controller moves.     

Nonlinear dimension reduction based neural modeling for distributed parameter processes

Many chemical processes are nonlinear distributed parameter systems with unknown uncertainties. For this class of infinite-dimensional systems, the low-order model identification from process data is very important in practice. The dimension reduction with a principal component analysis (PCA) is only a linear approximation for nonlinear problem. In this study, a nonlinear dimension reduction based low-order neural model identification approach is proposed for nonlinear distributed parameter processes. First, a nonlinear principal component analysis (NL-PCA) network is designed for the nonlinear dimension reduction, which can transform the high-dimensional spatio-temporal data into a low-dimensional time domain. Then, a neural system can be easily identified to model this low-dimensional temporal data. Finally, the spatio-temporal dynamics can be reproduced using the nonlinear time/space reconstruction. The simulations on a typical nonlinear transport-reaction process show that the proposed approach can achieve a better performance than the linear PCA based modeling approach.     

Distributed model predictive control of switched nonlinear systems with scheduled mode transitions

A method for the design of distributed model predictive control (DMPC) systems for a class of switched nonlinear systems for which the mode transitions take place according to a prescribed switching schedule is presented. Under appropriate stabilizability assumptions on the existence of a set of feedback controllers that can stabilize the closed‐loop switched, nonlinear system, a cooperative DMPC architecture using Lyapunov‐based model predictive control (MPC) in which the distributed controllers carry out their calculations in parallel and communicate in an iterative fashion to compute their control actions is designed. The proposed DMPC design is applied to a nonlinear chemical process network with scheduled mode transitions and its performance and computational efficiency properties in comparison to a centralized MPC architecture are evaluated through simulations. © 2013 American Institute of Chemical Engineers AIChE J, 59:860‐871, 2013     

基于复合神经网络的离散非线性系统建模与控制

An adaptive inverse controller for nonliear discrete-time system is proposed in this paper. A compound neural network is constructed to identify the nonlinear system, which includes a linear part to approximate the nonlinear system and a recurrent neural network to minimize the difference between the linear model and the real nonlinear system. Because the current control input is not included in the input vector of recurrent neural network (RNN), the inverse control law can be calculated directly. This scheme can be used in real-time nonlinear single-input single-output (SISO) and multi-input multi-output (MIMO) system control with less computation work. Simulation studies have shown that this scheme is simple and affects good control accuracy and robustness.     

Machine-learning-based construction of barrier functions and models for safe model predictive control

In this paper, we propose a control Lyapunov-barrier function-based model predictive control method utilizing a feed-forward neural network specified control barrier function (CBF) and a recurrent neural network (RNN) predictive model to stabilize nonlinear processes with input constraints, and to guarantee that safety requirements are met for all times. The nonlinear system is first modeled using RNN techniques, and a CBF is characterized by constructing a feed-forward neural network (FNN) model with unique structures and properties. The FNN model for the CBF is trained based on data samples collected from safe and unsafe operating regions, and the resulting FNN model is verified to demonstrate that the safety properties of the CBF are satisfied. Given sufficiently small bounded modeling errors for both the FNN and the RNN models, the proposed control system is able to guarantee closed-loop stability while preventing the closed-loop states from entering unsafe regions in state-space under sample-and-hold control action implementation. We provide the theoretical analysis for bounded unsafe sets in state-space, and demonstrate the effectiveness of the proposed control strategy using a nonlinear chemical process example with a bounded unsafe region.     

Machine learning-based distributed model predictive control of nonlinear processes

This work explores the design of distributed model predictive control (DMPC) systems for nonlinear processes using machine learning models to predict nonlinear dynamic behavior. Specifically, sequential and iterative DMPC systems are designed and analyzed with respect to closed-loop stability and performance properties. Extensive open-loop data within a desired operating region are used to develop long short-term memory (LSTM) recurrent neural network models with a sufficiently small modeling error from the actual nonlinear process model. Subsequently, these LSTM models are utilized in Lyapunov-based DMPC to achieve efficient real-time computation time while ensuring closed-loop state boundedness and convergence to the origin. Using a nonlinear chemical process network example, the simulation results demonstrate the improved computational efficiency when the process is operated under sequential and iterative DMPCs while the closed-loop performance is very close to the one of a centralized MPC system.     

The control of MSF desalination plants based on inverse model control by neural network

In this paper, a nonlinear inverse model control strategy based on neural network is proposed for MSF desalination plant. Artificial neural networks (ANNs) can handle complex and nonlinear process relationships, and are robust to noisy data. The designed neural networks consist of three layers identified from input–output data and trained with a descent gradient algorithm. The set point tracking performance of the proposed method was studied when the disturbance is present in the MSF system. Three controllers are designed for controlling the top brine temperature, the level of last stage and salinity. These results show that a neural network inverse model control strategy (NNINVMC) is robust and highly promising to be implemented in such nonlinear systems. Also the comparison between the top brine temperature of the proposed model and NN predicted data from the literature supports the accuracy of the model.     

基于ARX-NNPLS模型的优化策略及在聚合物牌号切换过程中的应用(英文)

Since it is often difficult to build differential algebraic equations (DAEs) for chemical processes, a new data-based modeling approach is proposed using ARX (AutoRegressive with eXogenous inputs) combined with neural network under partial least squares framework (ARX-NNPLS), in which less specific knowledge of the process is required but the input and output data. To represent the dynamic and nonlinear behavior of the process, the ARX combined with neural network is used in the partial least squares (PLS) inner model between input and output latent variables. In the proposed dynamic optimization strategy based on the ARX-NNPLS model, neither parameterization nor iterative solving process for DAEs is needed as the ARX-NNPLS model gives a proper representation for the dynamic behavior of the process, and the computing time is greatly reduced compared to conventional control vector parameterization method. To demonstrate the ARX-NNPLS model based optimization strategy, the polyethylene grade transition in gas phase fluidized-bed reactor is taken into account. The optimization results show that the final optimal trajectory of quality index determined by the new approach moves faster to the target values and the computing time is much less.     

NEURAL NETWORK-BASED PREDICTIVE CONTROL FOR MULTIVARIABLE PROCESSES

In this paper, the systematic derivations of setting up a nonlinear model predictive control based on the neural network are presented. This extends our previous work (Chen, 1998) into a multivariable system to explore the characteristics of the design. There are two stages for the development of nonlinear neural network predictive control: a neural network model and a control design. In the neural network model design, a parallel multiple-input, single-output neural network autoregressive with a model of exogenous inputs (NNARX) is proposed for multistep ahead predictions. In control design, the controller with extended control horizon is developed. The Levenberg-Marquardt algorithm is applied to training the NNARX model. Also, the sequential quadratic programming is used to search for the optimal manipulated inputs. The gradient of the objective function and constraints that require computation of Jacobian matrices are completely derived for optimization calculation. To demonstrate the control ability of MIMO cases, the proposed method is applied through two nonlinear simulation problems.     

Neural network-based predictive control for multivariable processes

In this paper, the systematic derivations of setting up a nonlinear model predictive control based on the neural network are presented. This extends our previous work (Chen, 1998) into a multivariable system to explore the characteristics of the design. There are two stages for the development of nonlinear neural network predictive control: a neural network model and a control design. In the neural network model design, a parallel multiple-input, single-output neural network autoregressive with a model of exogenous inputs (NNARX) is proposed for multistep ahead predictions. In control design, the controller with extended control horizon is developed. The Levenberg-Marquardt algorithm is applied to training the NNARX model. Also, the sequential quadratic programming is used to search for the optimal manipulated inputs. The gradient of the objective function and constraints that require computation of Jacobian matrices are completely derived for optimization calculation. To demonstrate the control ability of MIMO cases, the proposed method is applied through two nonlinear simulation problems.     

Grey-box modelling and control of chemical processes

The success of model based control of chemical processes is dependent on good process models. Many of these processes exhibit strong nonlinearity and time varying parameters and are often difficult to model accurately. The ‘grey-box’ model which combines partial knowledge of the process, with a neural network to capture the remaining dynamics, is a promising modelling tool for nonlinear processes. This modelling methodology maximizes the use of a priori process knowledge. This, in turn, reduces the size of the neural network required to capture the remaining dynamics, hence, less data for training and faster convergence can be achieved. The grey-box model is combined with a generic model control structure and applied to a number of simulations as well as a real-time process.     

A simple adaptive control strategy for temperature trajectory tracking in batch processes

In this paper, a simple adaptive control strategy is suggested for temperature tracking control of batch processes. A nonlinear controller, which is in structure very simple and consists of a single parameter, is proposed. To enable this controller to control a batch process adaptively, a simple parameter tuning algorithm is derived based on the Lyapunov stability theorem. The proposed adaptive control scheme is directly operational, which does not depend on process model and the only a priori process information required is the system response direction. To demonstrate the effectiveness and applicability of the proposed scheme, illustrative examples are provided. Extensive simulation results reveal that the proposed adaptive control strategy appears to be a simple and effective approach to batch process control, which provides robust control despite the wide range of operating conditions and nonlinear dynamics of the system.     

基于稳定Hammerstein模型的在线软测量建模方法及应用

针对复杂工业过程中由于存在未建模动态和不确定干扰,导致关键变量的软测量精度下降的问题,提出了一种基于稳定Hammerstein模型(H模型)的在线软测量建模方法。H模型的非线性增益采用带有时变稳定学习算法的小波神经网络模型,线性系统部分采用基于递推最小二乘的ARX模型,基于输入到状态稳定性理论证明了H模型辨识误差的有界性。其中小波神经网络具有表征强非线性的特性,稳定学习算法可抑制未建模动态和不确定干扰的影响,改善了模型的预测精度和自适应能力。以典型非线性系统和实际污水处理过程为例进行了仿真研究,结果表明,基于稳定H模型的软测量方法具有较高的在线软测量精度。     

Modelling and control of different types of polymerization processes using neural networks technique: A review

Polymerization process can be classified as a nonlinear type process since it exhibits a dynamic behaviour throughout the process. Therefore, it is highly complicated to obtain an accurate mechanistic model from the nonlinear process. This predicament always been a “wall” to researchers to be able to devise an optimal process model and control scheme for such a system. Neural networks have succeeded the other modelling and control methods especially in coping with nonlinear process due to their very conciliate characteristics. These characteristics are further explained in this work. The predicament that is encountered by researchers nowadays is lack of data which consequently lead to an imprecise mechanistic model that scarcely conforms to the desired process. The implementations of the neural network model not only restrict to polymerization reactor but to other difficult‐to‐measure parameters such as polymer quality, polymer melts index and mixture of initiators. This work is aimed to manifest ascendancy of neural networks in modelling and control of polymerization process.     

Generic model controller tuning for chemical processes with input saturation

Control in the face of process input constraints is very common and of great practical importance in the processing industries. Generic Model Control (GMC) is a model‐based control framework for both linear and nonlinear systems. In this paper, a constrained GMC controller tuning approach using a nonlinear least squares technique is proposed. This tuning approach is simple to apply. For a SISO GMC control system with input saturation, the tracking performance is significantly improved by adding a simple heuristic switching strategy. The effectiveness of the proposed controller tuning approach is demonstrated using dynamic simulations and MIMO real‐time experiments.     

Online machine learning modeling and predictive control of nonlinear systems with scheduled mode transitions

This work develops a model predictive control (MPC) scheme using online learning of recurrent neural network (RNN) models for nonlinear systems switched between multiple operating regions following a prescribed switching schedule. Specifically, an RNN model is initially developed offline to model process dynamics using the historical operational data collected in a small region around a certain steady-state. After the system is switched to another operating region under a Lyapunov-based MPC with suitable constraints to ensure satisfaction of the prescribed switching schedule policy, RNN models are updated using real-time process data to improve closed-loop performance. A generalization error bound is derived for the updated RNN models using the notion of regret, and closed-loop stability results are established for the switched nonlinear system under RNN-based MPC. Finally, a chemical process example with the operation schedule that requires switching between two steady-states is used to demonstrate the effectiveness of the proposed RNN-MPC scheme.     

On‐Line PEMFC Control Using Parameterized Nonlinear Model‐Based Predictive Control

In this work, a fast nonlinear model‐based predictive control (NMPC) strategy is designed and experimentally validated on‐line on a real fuel cell. Regarding NMPC strategies, the most challenging part remains to achieve on‐line implementation, especially when dealing with fast dynamic systems. As previously demonstrated in a recent work, the proposed control strategy is ideally suited to address this problem. Indeed, it is 30 times faster than classical NMPC controllers. This strategy relies on a specific parameterization of the control actions to reduce the computational time and achieve on‐line implementation. Due to its short computational time compared to mechanistic models, an artificial neural network model is designed and experimentally validated. This model is employed as internal model in the NMPC controller to predict the system behavior. To confirm the applicability and the relevance of the proposed NMPC controller varying control scenarios are investigated on a test bench. The built‐in controller is overridden and the NMPC controller is implemented externally and executed on‐line. Experimental results exhibit the outstanding tracking capability and robustness against model‐process mismatch of the proposed strategy. The parameterized NMPC controller turns out to be an excellent candidate for on‐line applications.     

基于自适应模糊神经网络的非线性系统模型预测控制

针对非线性动态系统的控制问题,提出了一种基于自适应模糊神经网络(adaptive fuzzy neural network, AFNN)的模型预测控制(model predictive control, MPC)方法。首先,在离线建模阶段,AFNN采用规则自分裂技术产生初始模糊规则,采用改进的自适应LM学习算法优化网络参数;然后,在实时控制过程,AFNN根据系统输出和预测输出之间的误差调整网络参数,从而为MPC提供一个精确的预测模型;进一步,AFNN-MPC利用带有自适应学习率的梯度下降寻优算法求解优化问题,在线获取非线性控制量,并将其作用到动态系统实施控制。此外,给出了AFNN-MPC的收敛性和稳定性证明,以保证其在实际工程中的成功应用。最后,利用数值仿真和双CSTR过程进行实验验证。结果表明,AFNN-MPC能够取得优越的控制性能。     

Development of a hybrid neural network model to predict feeding method in fed‐batch cultivation for enhanced recombinant streptokinase productivity in Escherichia coli

BACKGROUND: A simple and efficient model for enhancing production of recombinant proteins is essential for cost effective development of processes at industrial scale. A hybrid neural network (HNN) model is proposed combining an unstructured model and neural network to predict the feeding method for the post‐induction phase of fed‐batch cultivation for increased recombinant streptokinase activity in Escherichia coli. RESULTS: The parameters of the unstructured model were estimated from experiments conducted with various feeding methods. The simulated model described the dynamics of the process satisfactorily, however, its predictive capability of the process for different feeding methods is limited due to wide disparity in process parameters. In contrast, a neural network model trained to map the variations in process parameters to state variables complements the ‘first principle’ model in predicting the state variables effectively. CONCLUSIONS: The HNN model is able to predict the product profile for different substrate feed rates. Further, the average volumetric streptokinase activity predicted by the HNN model matches closely the experimental values for fed‐batches having high as well as low streptokinase activity. The HNN model developed in this study could facilitate development of a process for recombinant protein production with minimum number of experiments. Copyright © 2011 Society of Chemical Industry     

Flexible on‐line Modeling and Control of pH in Waste Neutralization Reactors

The control of pH in waste neutralization processes presents a challenging highly nonlinear and time‐varying problem in which the reactor also suffers from inaccessible state information. The ability to characterize the changing dynamics of such reactors is essential to the success of advanced control schemes for these applications. In this work, flexible on‐line modeling of a pH reactor simulating nonstationary behavior was studied. This entailed a comparison of the most popular connectionist learning algorithm, the “Widrow‐Hoff delta rule”, with a classical tool in adaptive identification and control, recursive least squares (RLS). The modeling was pursued within the framework of neural networks using the ADALINE neural network. Further, two heuristically defined first‐principles‐based transforms were investigated for providing “general globally linearizing” information to the ADALINE. The comparisons of the learning algorithms for different neural network information vectors has led to a critical understanding of the flexibility of each algorithm for on‐line learning of the diverse process gain characteristics encountered in pH reactors.