Batch-to-batch optimal control of a batch polymerisation process based on stacked neural network models

A neural network based batch-to-batch optimal control strategy is proposed in this paper. In order to overcome the difficulty in developing mechanistic models for batch processes, stacked neural network models are developed from process operational data. Stacked neural networks have enhanced model generalisation capability and can also provide model prediction confidence bounds. However, the optimal control policy calculated based on a neural network model may not be optimal when applied to the true process due to model plant mismatches and the presence of unknown disturbances. Due to the repetitive nature of batch processes, it is possible to improve the operation of the next batch using the information of the current and previous batch runs. A batch-to-batch optimal control strategy based on the linearisation of stacked neural network model is proposed in this paper. Applications to a simulated batch polymerisation reactor demonstrate that the proposed method can improve process performance from batch to batch in the presence of model plant mismatches and unknown disturbances.     

Multi-objective optimisation of batch separation processes

This paper considers for the first time the simultaneous multi-objective optimisation of design and operation of batch distillation as well as of batch hybrid distillation/pervaporation processes. The overall problem is formulated as a multi-objective mixed integer dynamic optimisation (MO-MIDO) problem. The optimisation strategy comprises of different ranking procedures that allow the determination of the Pareto optimal set. A case study for the separation of a homogeneous tangent-pinch (acetone–water) mixture is presented for a dual-criteria optimisation case of minimising capital investment while at the same time minimising the energy consumption rate during the batch. It is found that the proposed distance ranking procedure yields the best Pareto optimal set when compared to other non-dominated sorting procedures. Furthermore, the distance ranking procedure was found to be further improved when used with an elitism operator.     

Parameter estimation in batch process using EM algorithm with particle filter

This paper investigates a parameter estimation problem for batch processes through the maximum likelihood method. In batch processes, the initial state usually relates to the states of previous batches. The proposed algorithm takes batch-to-batch correlations into account by employing an initial state transition equation to model the dynamics along the batch dimension. By treating the unmeasured states and the parameters as hidden variables, the maximum likelihood estimation is accomplished through the expectation–maximization (EM) algorithm, where the smoothing for the terminal state and the filtering for the initial state are specially considered. Due to the nonlinearity and non-Gaussianity in the state space model, particle filtering methods are employed for the implementation of filtering and smoothing. Through alternating between the expectation step and the maximization step, the unknown parameters along with states are estimated. Simulation examples demonstrate the proposed estimation approach.     

Emission control in palm oil mills using artificial neural network and genetic algorithm

The present study utilized a combination of artificial neural network (ANN) and genetic algorithms (GA) to optimize the release of emission from the palm oil mill. A model based on ANN is developed from the actual data taken from the palm oil mill. The predicted data agree well with the actual data taken. GA is then employed to find the optimal operating conditions so that the overlimit release of emission is reduced to the allowable limit.     

Application of recurrent neural networks in batch reactors: Part I. NARMA modelling of the dynamic behaviour of the heat transfer fluid temperature

This paper is focused on the development of nonlinear models, using artificial neural networks, able to provide appropriate predictions when acting as process simulators. The dynamic behaviour of the heat transfer fluid temperature in a jacketed chemical reactor has been selected as a case study. Different structures of NARMA (Non-linear ARMA) models have been studied. The experimental results have allowed to carry out a comparison between the different neural approaches and a first-principles model. The best neural results are obtained using a parallel model structure based on a recurrent neural network architecture, which guarantees better dynamic approximations than currently employed neural models. The results suggest that parallel models built up with recurrent networks can be seen as an alternative to phenomenological models for simulating the dynamic behaviour of the heating/cooling circuits which change from batch installation to installation.     

基于结构逼近式神经网络的间歇反应器优化控制

利用结构逼近式混合神经网络（SAHNN）建立了一类典型放热液相二级平行间歇反应的数学模型。基于主产物浓度和反应温度的递归神经网络（RNN）模型,使用混合PSO-SQP算法求解该间歇反应主产物产率最大化问题,进而得到反应温度优化曲线。鉴于反应温度实时可测,提出扩展的EISE指标,该指标把实时计算的模型误差引入控制策略,为基于模型的控制增加了反馈通道,增强了控制方法的鲁棒性和抗干扰性能。利用 原理对所提出的一步超前预测控制做了稳定性分析,证明了算法的正确性。研究的结果充分证明了基于SAHNN混合神经网络模型的优化控制策略的有效性。     

Optimal control of diffusion-convection-reaction processes using reduced-order models

Two approaches for optimal control of diffusion-convection-reaction processes based on reduced-order models are presented. The approaches differ in the way spatial discretization is carried out to compute a reduced-order model suitable for controller design. In the first approach, the partial differential equation (PDE) that describes the process is first discretized in space and time using the finite difference method to derive a large number of recursive algebraic equations, which are written in the form of a discrete-time state-space model with sparse state, input and output matrices. Snapshots based on this high-dimensional state-space model are generated to calculate empirical eigenfunctions using proper orthogonal decomposition. The Galerkin projection with the computed empirical eigenfunctions as basis functions is then directly applied to the high-dimensional state-space model to derive a reduced-order model. In the second approach, a continuous-time finite-dimensional state-space model is constructed directly from the PDE through application of orthogonal collocation on finite elements in the spatial domain. The dimension of the derived state-space model can be further reduced using standard model reduction techniques. In both cases, optimal controllers are designed based on the low-order state-space models using discrete-time and continuous-time linear quadratic regulator (LQR) techniques. The effectiveness of the proposed methods are illustrated through applications to a diffusion-convection process and a diffusion-convection-reaction process.     

Coatings of Na-aluminosilicate zeolites prepared using predictions from an artificial neural network method

Predictions made by the Generalized Regression Neural Networks (GRNN) method were used to relate the initial compositions of various reaction mixtures to the types of Na-aluminosilicate zeolites that may be obtained from these compositions. In the light of the predictions made, coatings were prepared on stainless steel plates, which were characterized by X-ray diffraction and scanning electron microscopy prior to and after syntheses. Coatings of zeolites P, X, A, analcime as well as their mixtures could be obtained from a variety of previously unknown clear solution compositions, generally in good accordance with the predictions made by the GRNN method. Different textural properties were obtained for the coatings of the same zeolite, such as P and X, which could be prepared from a relatively wide range of compositions.     

Iterative learning control for the systematic design of supersaturation controlled batch cooling crystallisation processes

The paper presents an approach to improve the product quality from batch-to-batch by exploiting the repetitive nature of batch processes to update the operating trajectories using process knowledge obtained from previous runs. The data based methodology is focused on using the linear time varying (LTV) perturbation model in an iterative learning control (ILC) framework to provide a convergent batch-to-batch improvement of the process performance indicator. The major contribution of this work is the development of a novel hierarchical ILC (HILC) scheme for systematic design of the supersaturation controller (SSC) of seeded batch cooling crystallizers. The HILC is used to determine the required supersaturation setpoint for the SSC and the corresponding temperature trajectory required to produce crystals with desired end-point property. The performance and robustness of these approaches are evaluated through simulation case studies. These results demonstrate the potential of the ILC approaches for controlling batch processes without rigorous process models.     

Joint analysis of transient flux behaviors via membrane fouling in hybrid PAC/MF processes using neural network

The main purpose of this study is to understand the transient flux behaviors via membrane fouling in hybrid powdered activated carbon/microfiltration (PAC/MF) processes. Experiments were carried out for nickel solution at various surfactant adsorbent and membrane types and membrane pore sizes. Transient fluxes were modeled together using neural network (NN). A good agreement was obtained with correlation value of 0.986 and mean absolute error of 0.366 m³/m² h respectively. In hybrid processes three discriminable phase behaviors for the flux were determined as fast [0-40s] slow [40-1000s] and low stable [1000-4500s]. Considerable losses in the flux were determined during the first and the second phases while the flux reached a partial steady-state in the third phase. In the first and the third phases participation of surfactant-adsorbed PACs to the cake layer and/or interaction of surfactants and PACs in the cake with each other were found to be predominant ways for the fouling. Whereas in the second phase the interaction of surfactants with both membrane and cake layer was appeared more influential on transient flux behavior.     

Robustness of reinforced gradient-type iterative learning control for batch processes with Gaussian noise☆

In this paper, a reinforced gradient-type iterative learning control profile is proposed by making use of system matrices and a proper learning step to improve the tracking performance of batch processes disturbed by exter-nal Gaussian white noise. The robustness is analyzed and the range of the step is specified by means of statistical technique and matrix theory. Compared with the conventional one, the proposed algorithm is more efficient to resist external noise. Numerical simulations of an injection molding process il ustrate that the proposed scheme is feasible and effective.     

Iterative learning reliable control of batch processes with sensor faults

An iterative learning reliable control (ILRC) scheme is developed in this paper for batch processes with unknown disturbances and sensor faults. The batch process is transformed into and treated as a two-dimensional Fornasini-Marchesini (2D-FM) model. Under the proposed control law, the closed-loop system with unknown disturbances and sensor faults not only converges along both the time and the cycle directions, but also satisfies certain H_∞ performance. For performance comparison, a traditional reliable control (TRC) law based on dynamic output feedback is also developed by considering the batch process in each cycle as a continuous process. Conditions for the existence of ILRC scheme are given as biaffine and linear matrix inequalities. Algorithms are given to solve these matrix inequalities and to optimize performance indices. Applications to injection packing pressure control show that the proposed scheme can achieve the design objectives well, with performance improvement along both time and cycle directions, and also has good robustness to uncertain initialization and measurement disturbances.     

Evaluation of optimisation techniques and control variable formulations for a batch cooling crystallization process

The mathematical optimisation of a batch cooling crystallization process is considered in this work. The objective is to minimize the standard deviation of the final crystal size distribution (CSD), which is an important feature in many industrial processes. The results with the problem written as a nonlinear programming and solved with the successive quadratic programming (SQP) coupled with the discretization of the control variable are compared with those obtained when SQP coupled with the parameterisation of the control variable is applied. Also it is proposed the implementation of the genetic algorithm (GA) coupled with parameterisation of the control variable. Extensive evaluations show that the SQP method is sensitive both to the parameterisation formulation and to the initial estimate. The solution with GA provided the control variable profile that leads to the minimum standard deviation of the final CSD. Nevertheless, it is a very time-consuming technique, which hampers its utilization in real time applications. However, its feature of global searching suggests its suitability in solving offline problems, in order to provide initial setup profiles. Bearing this in mind, it is proposed an algorithm which allows for the implementation of GA solution in a real time fashion, taking advantage of its robustness to find out the optimal solution.     

Sustainable scheduling of batch processes under economic and environmental criteria with MINLP models and algorithms

We address the bi-criterion optimization of batch scheduling problems with economic and environmental concerns. The economic objective is expressed in terms of productivity, which is the profit rate with respect to the makespan. The environmental objective is evaluated by means of environmental impact per functional unit based on the life cycle assessment methodology. The bi-criterion optimization model is solved with the ε-constraint method. Each instance is formulated as a mixed-integer linear fractional program (MILFP), which is a special class of non-convex mixed-integer nonlinear programs. In order to globally optimize the resulting MILFPs effectively, we employ the tailored reformulation-linearization method and Dinkelbach's algorithm. The optimal solutions lead to a Pareto frontier that reveals the tradeoff between productivity and environmental impact per functional unit. To illustrate the application, we present two case studies on the short-term scheduling of multiproduct and multipurpose batch plants.     

A neural network-based soft sensor for particle size distribution using image analysis

Many industrial processes require on-line measurement of particle size and particle size distribution for process monitoring and control. The available techniques for reliable on-line measurement are, however, limited. In this paper, based on the captured surface images of randomly disarranged ore particles, the image uniformity was characterized. Particle size distribution was then investigated by applying a neural network-based modeling with the obtained image uniformity. The proposed soft sensor provides an improved prediction model and can be used for real time measurement of particle size distribution in the industrial operations.     

Asynchronous optimisation with the use of a cascade search algorithm

This paper introduces the development of an asynchronous approach coupled with a cascade optimisation algorithm. The approach incorporates concepts of asynchronous Markov processes and introduces a search process that is benefiting from distributed computing infrastructures. The algorithm uses concepts of partitions and pools to store intermediate solutions and corresponding objectives. Population inflections are performed periodically to ensure that Markov processes, still independent and asynchronous, make arbitrary use of intermediate solutions. Tested against complex optimisation problems and in comparison with commonly used Tabu Search, the asynchronous cascade algorithm demonstrates a significant potential in distributed operations with favourable comparisons drawn against synchronous and quasi-asynchronous versions of conventional algorithms.     

Process analysis and optimisation of hybrid processes for the dehydration of ethanol

Promising hybrid processes for ethanol dewatering consist of different combinations of distillation with adsorption and/or vapour permeation. This paper presents an analysis and optimisation of these hybrid processes using non-equilibrium models and an evolutionary algorithm. Four different membrane assisted configurations are compared with a benchmark process consisting of distillation and pressure swing adsorption. In total 12 cases were investigated while assuming different feed and product compositions at different production capacities: three ethanol mass fractions in feed 45, 80, 92 wt.%, two product purities 99.6, 99.95 wt.% and two production capacities 25,000, 250,000 m³/year. The influence of decisive operating and structural variables on important target variables such as total membrane area is demonstrated. Finally, the processes are evaluated regarding operating costs and energy consumption depending on product purity and production capacity. The operating costs of the membrane assisted configurations differ only in a small range of −3% to 6% from those of the benchmark. The energy consumption of the membrane assisted configurations without distillation is up to 30% lower compared to the benchmark. Especially the combination of vapour permeation and adsorption is a promising alternative allowing for producing ethanol with high purities at lower operating pressures compared to the vapour permeation as stand alone process.     

A neural network approach to the prediction of diesel fuel lubricity

The continuous sulfur reduction in diesel fuel has resulted in poor fuel lubricity and engine pump failure, a fact that led to the development of a number of methods that measure the actual fuel lubricity level. However, lubricity measurement is costly and time consuming, and a number of predictive models have been developed in the past, based mainly on various fuel properties. In the present paper, a black box modeling approach is proposed, where the lubricity is approximated by a radial basis function (RBF) neural network that uses other fuel properties as inputs. The HFRR apparatus was used for lubricity measurements. In the present model, the variables used included the diesel fuel conductivity, density, kinematic viscosity at 40 °C, sulfur content and 90% distillation point, which produced the smallest error in the validation data.     

Adaptive multiscale convolutional neural network model for chemical process fault diagnosis

Intelligent fault recognition techniques are essential to ensure the long-term reliability of manufacturing.Due to the variations in material,equipment and environment,the process variables monitored by sensors contain diverse data characteristics at different time scales or in multiple operating modes.Despite much progress in statistical learning and deep learning for fault recognition,most models are constrained by abundant diagnostic expertise,inefficient multiscale feature extraction and unr...     

Dynamic hybrid simulation of batch processes driven by a scheduling module

Simulation is now a CAPE tool widely used by practicing engineers for process design and control. In particular, it allows various offline analyses to improve system performance such as productivity, energy efficiency, waste reduction, etc. In this framework, we have developed the dynamic hybrid simulation environment PrODHyS whose particularity is to provide general and reusable object-oriented components dedicated to the modeling of devices and operations found in chemical processes. Unlike continuous processes, the dynamic simulation of batch processes requires the execution of control recipes to achieve a set of production orders. For these reasons, PrODHyS is coupled to a scheduling module (ProSched) based on a MILP mathematical model in order to initialize various operational parameters and to ensure a proper completion of the simulation. This paper focuses on the procedure used to generate the simulation model corresponding to the realization of a scenario described through a particular scheduling.