多节点流水线生产订单接受与调度一体化问题

为了使企业在选择订单时获取最大利润,避免产生惩罚损失,提出了一种流水线生产企业订单接受与调度一体化的决策方法。在该方法中,将订单接受与调度同时规划,建立了以利润最大化为目的,考虑拖期惩罚的决策模型。提出一种新的基于模拟退火过程的启发式求解算法来求解该模型,实例验证了算法的有效性。     

Reinforcement learning for batch process control: Review and perspectives

Batch or semi-batch processing is becoming more prevalent in industrial chemical manufacturing but it has not benefited from advanced control technologies to a same degree as continuous processing. This is due to its several unique aspects which pose challenges to implementing model-based optimal control, such as its highly nonstationary operation and significant run-to-run variabilities. While existing advanced control methods like model predictive control (MPC) have been extended to address some of the challenges, they still suffer from certain limitations which have prevented their widespread industrial adoption. Reinforcement learning (RL) where the agent learns the optimal policy by interacting with the system offers an alternative to the existing model-based methods and has potential for bringing significant improvements to industrial batch process control practice. With such motivation, this paper examines the advantages that RL offers over the traditional model-based optimal control methods and how it can be tailored to better address the characteristics of industrial batch process control problems. After a brief review of the existing batch control methods, the basic concepts and algorithms of RL are introduced and issues for applying them to batch process control problems are discussed. The nascent literature on the use of RL in batch process control is briefly reviewed, both in recipe optimization and tracking control, and our perspectives on future research directions are shared.     

Robust controller design for temperature tracking problems in jacketed batch reactors

There is an increasing trend to employ advanced instrumentation and control strategies for batch processes where expensive products are being manufactured. In this paper, a robust nonlinear control strategy is developed for temperature tracking problems in batch reactors in the presence of parametric uncertainty. The controller has a multi-loop feedback configuration. An inner loop is designed for approximate input–output linearization of a nominal plant. The outer loop is designed for stability and robust performance by utilizing results from structured singular values (μ-synthesis). It is shown via simulation of a temperature tracking problem in batch synthesis that the controller provides excellent tracking despite parametric uncertainty.     

基于平均强化学习的订单生产方式企业订单接受策略

从收益管理思想出发,采用平均强化学习算法研究不确定环境下订单生产(MTO)方式企业的订单接受问题。以最大化平均期望收益为优化目标,采用多级价格机制,把订单类型、价格和提前期的不同组合作为系统状态划分标准,结合平均强化学习原理,提出了具有学习能力的订单接受算法(RLOA)。仿真结果表明,RLOA算法具有学习和选择性接受订单的能力,与其他订单接受规则相比,在平均收益、订单类型接受状况和适应性等方面都有较好表现。     

A feedback-based implementation scheme for batch process optimization

The terminal-cost optimization of a control–affine nonlinear system leads to a discontinuous solution that can be characterized in a piecewise manner. To implement such an optimal trajectory despite disturbances and parametric uncertainty, a cascade optimization scheme is proposed in this paper, where optimal reference signals are tracked. Optimality is achieved by the appropriate definition of reference signals (input bounds, state constraints, or switching functions) to track in various sub-intervals. Furthermore, conservatism is introduced into the optimization problem to ensure satisfaction of path constraints in the presence of uncertainty. Finally, the proposed cascade optimization scheme is illustrated on a simulation of a fed-batch penicillin fermentation plant.     

An optimal batch size for a JIT manufacturing system

This paper addresses the problem of a manufacturing system that procures raw materials from suppliers in a lot and processes them to convert to finished products. It proposes an ordering policy for raw materials to meet the requirements of a production facility. In turn, this facility must deliver finished products demanded by outside buyers at fixed interval points in time. In this paper, first we estimate production batch sizes for a JIT delivery system and then we incorporate a JIT raw material supply system. A simple algorithm is developed to compute the batch sizes for both manufacturing and raw material purchasing policies. Computational experiences of the problem are also briefly discussed.     

Evaluation of neural networks-based controllers in batch polymerisation of methyl methacrylate

The importance of batch reactors in today's process industries cannot be overstated. Thus said, it is important to optimise their operation in order to consistently achieve products of high quality while minimising the production of undesirables. In processes like polymerisation, these reactors are responsible for a greater number of products than other reactor types and the need for optimal operation is therefore greater.

An approach based on an offline dynamic optimisation and online control strategy is used in this work to generate optimal set point profiles for the batch polymerisation of methyl methacrylate. Dynamic optimisation is carried out from which controller set points to attain desired polymer molecular end point characteristics are achieved. Temperature is the main variable to be controlled, and this is done over finite discrete intervals of time.

For on-line control, we evaluate the performance of neural networks in two controllers used to track the derived optimal set points for the system. The controllers are generic model control (GMC), ([P.L. Lee, G.R. Sullivan, Generic model control, Comput. Chem. Eng. 12(6) (1998) 573–580]) and the neural network-based inverse model-based control (IMBC), ([M.A. Hussain, L.S. Kershenbaum, Implementation of an inverse model based control strategy using neural networks on a partially simulated exothermic reactor, Trans. IchemE 78(A) (2000) 299–311]). Although the GMC is a model-based controller, neural networks are used to estimate the heat release within its framework for on-line control. Despite the application of these two controllers to general batch reactors, no published work exists on their application to batch polymerisation in the literature. In this work, the performance of the neural networks within each controller's algorithm for tracking and setpoint regulation of the optimal trajectory and in robustness tests on the system is evaluated.     

Iterative learning control-based batch process control technique for integrated control of end product properties and transient profiles of process variables

Importance of batch processes has grown recently with the increasing economic competition that has pushed the manufacturing industries to pursue small quantity production of diverse high value-added products. Accordingly, systems engineering research on advanced control and optimization of batch processes has proliferated. In this paper, we examine the potentials of ‘iterative learning control (ILC)’ as a framework for industrial batch process control and optimization. First, various ILC rules are reviewed to provide a historical perspective. Next it is shown how the concept of ILC can be fused with model predictive control (MPC) to build an integrated end product and transient profile control technique for industrial chemical batch processes. Possible extensions and modifications of the technique are also presented along with some numerical illustrations. Finally, other related techniques are introduced to note the similarities and contemplate the opportunities for synergistic integration with the current ILC framework.     

Impact of automated decision aids on performance,operator behaviour and workload in a simulated supervisory control task

In studies reporting automation effects on overall system performance and on the operator, the methods used to measure workload often did not appropriately reflect the complexity of this construct. The present study addresses the impact of automation on operator workload and behaviour in process control fault management. Workload effects were assessed with subjective, cardiovascular and secondary task performance indicators. Interactions with the interface of the process control simulation directed at gathering information and controlling the system were recorded. Automation made operators more efficient, allowing faster fault management with less information sampling and control actions. Subjective workload ratings were significantly lower in the assisted conditions as compared to manual, which was not reflected in cardiovascular and secondary task measures. Participants' information sampling activity did not differ between medium and high level of automation. Results suggest that participants paid constantly high attention to their task even with highly automated support.     

Effects of control order,augmented feedback,input device and practice on tracking performance and perceived workload

Virtual interfaces to advanced human-machine systems will present operators with a variety of perceptual-motor challenges. To inform the virtual interface design processes, the present experiments examined the effects of track order, level of knowledge of performance, type of control device, and the extent of practice on tracking performance and associated mental workload. Tracking was assessed by root mean square error. Subjective workload was measured using both the NASA Task Load Index (TLX) and the Subjective Workload Assessment Technique (SWAT). Results indicated non-linear effects, where tracking error and subjective workload both increased non-proportionally with track order. Trackball use resulted in more accurate performance and was judged to be of lower subjective workload than input using a mouse. Augmented knowledge of performance had little effect on either performance or workload. There were a number of interactions affecting performance that were replicated in perceived workload. Over acquisition trials, second-order tracking exhibited continuous improvement. This capability was retained even after a 30-day rest interval. Decrease in workload followed performance improvement in both initial acquisition and subsequent retention phases. The two subjective workload scales were essentially equivalent in terms of their sensitivity to task manipulations. These results support the direct association between workload and performance and confirms the use of workload in helping to evaluate the influence of diverse task-related demands. The implications for the design of virtual interfaces to real-world systems are examined in the light of these findings.     

Evaluating the impact of alternative plans on manufacturing performance

In recent years, process planners have become interested in the development of dynamic process planning systems that can interface to scheduling systems providing alternative process plans to increase flexibility in scheduling. However, deciding how many alternatives are needed has not been addressed in any previous studies. This paper presents the results of a simulation-based study aimed at characterizing the benefit provided from having alternative plans available for use in scheduling. This benefit is quantified in terms of the overall performance of a job-shop manufacturing environment. The results of this study indicate that the advantage gained by increasing the number of alternative process plans diminishes rapidly. In fact, under some conditions for the particular system studied, increasing the number of alternatives actually resulted in degraded system performance. Based on these results developers of process planning systems and methodologies need to evaluate carefully the benefit of expending time and resources on the generation of alternative plans or optimal plans.     

Improved monitoring of batch processes by incorporating external information

In this paper an overview is given of statistical process monitoring with the emphasis on batch processes and the possible steps to take for improving this by incorporating external information. First, the general concept of statistical process monitoring of batches is explained. This concept has already been shown to be successful according to the number of references to industrial applications. The performance of statistical process monitoring of batch processes can be enhanced by incorporating external information. Two types of external information can be distinguished: batch-run specific and process specific information. Various examples of both types of external information are given. Several ideas of how to incorporate the external information in model development are discussed. The concept of incorporating process specific information is highlighted by an example of a grey model. This model is applied to a biochemical batch process that is spectroscopically monitored.     

Fault prognosis for batch production based on percentile measure and gamma process: Application to semiconductor manufacturing

Batch manufacturing processes (BMP) play an important role in many production industries, such as in semiconductor, electronic and pharmaceutical industries. They generally exhibit some batch-to-batch or unit-to-unit variations due to many reasons such as variations in impurities and deviations of the process variables from their trajectories. The process monitoring for these systems has been considered as rather fault diagnosis than as fault prognosis, this latter has received scarce attention in the literature. This paper presents a data-driven prognostic method for BMP organized in three steps. The first step allows to reduce the data size and to extract a raw health index which represents the operating state of the system. In the second step, variations in the health index are processed by the percentile measure which is use in a way that gives rise to monotonic profiles. In the third step, these profiles are modelled by gamma process as it is the most appropriate for the stochastic modelling of monotonic and gradual deterioration. The remaining useful life (RUL) is then estimated using an aggregate probability density function (pdf) with a confidence interval (CI) that ensures the safety margins in industry. Finally, the proposed method is applied on semiconductor manufacturing equipment with two industrial datasets provided by STMicroelectronics.     

Trajectory optimization and positioning control for batch process using learning control

Efficiency and accuracy are critical in the motion control of a batch process. This paper proposes a new intelligent motion control method for a batch process based on reinforcement learning (RL) and iterative learning control (ILC). The proposed learning-based motion control method enables the system to learn from its previous experience. The motion control method can be divided into two parts: (1) RL-based trajectory optimization and (2) ILC-based positioning control. Experiments were conducted to demonstrate the effectiveness of the proposed method. The results indicate that the proposed method not only reduces the process time effectively while ensuring system stability, but also achieves excellent positioning accuracy.     

On the use of a synthetic online/batch workload for computer selection

The overall evaluation and decision process for the selection of a new computer system for the universities at Graz using a mixed synthetic online/batch workload is described. In particular the method for defining this workload for the evaluation process is explained. Furthermore the possibility of constructing assessment functions, that can be used for the comparison of different systems, is discussed.     

Latent variable MPC for trajectory tracking in batch processes

A novel multivariate empirical model predictive control strategy (LV-MPC) for trajectory tracking and disturbance rejection for batch processes is presented. The strategy is based on dynamic principal component analysis (PCA) models of the batch process. The solution to the control problem is computed in the low dimensional latent variable space of the PCA model. The trajectories of all variables over the future horizon are then computed from the latent variable solution of the controller. The excellent control performance and the modest closed-loop data requirements for identification are illustrated for the temperature tracking in simulations of an emulsion polymerization process, an exothermic chemical reaction system and for MIMO temperature and pressure tracking in a nylon polymerization autoclave.     

The role of cue utilisation in reducing the workload in a train control task

Abstract

Skilled performance has been characterised, in part, by the capacity to accurately identify and respond to patterns as cues in the environment. The outcome is a reduction in cognitive load and a greater residual capacity to undertake concurrent tasks. The present study was designed to examine the relationship between cue utilisation and temporal pattern recognition in the context of a simulated, rail control task. Sixty-one university students undertook an assessment of cue utilisation and engaged in a rail control simulation. The appearance and movement of trains followed a consistent but implicit (undisclosed) pattern. Throughout the second half of the rail task, a secondary task was included. The results indicated that participants with relatively higher cue utilisation were more likely to identify the implicit pattern of rail movements, were more accurate and responded more rapidly under increased workload conditions. The results suggest that a propensity to identify patterns as cues may provide an opportunity to reduce cognitive demands, thereby facilitating performance in a novel task. Implications for selection and system design are discussed.

Practitioner Summary: This study was designed to explain differences in the way in which people learn, particularly when tasks involve recurring patterns. Using simulated rail control, the results indicated that participants who display behaviour that is indicative of the utilisation of cues also recognise patterns in the movement of simulated trains. This enables them to manage trains more effectively, even while undertaking other tasks.     

The use of intelligent feedback for work order release in an uncertain manufacturing system

A turbulent manufacturing environment where uncertainty is inevitable does not allow for the availability of the required materials and resources when they are needed. This paper studies the implications of demand surges, lead-time variations and resources breakdown on the ability of a manufacturing system to achieve its delivery target. Simulation modelling was used to represent a stochastic manufacturing system, which is disturbed by these uncertainties. Manufacturing systems each with and without intelligent feedback were modelled. An intelligent feedback is represented via a set of algorithm, which reanalyse and self-organise the new status of the order in the presence of the uncertainties and update the relevant attributes before the order is released. Four types of intelligence were examined: (1) lead-time allowance, (2) capacity allowance, (3) safety stock allowance and (4) batching flexibility. Experiments results from each system were compared. It was found that the manufacturing system with intelligent feedback has a higher ability to achieve its delivery target by proactively tackling the variations caused by uncertainties. This study also found that the reliability of a work order in the presence of uncertainty could be improved by using an appropriate type of intelligence, which is dependent upon how and when the order was released. It was concluded from this research that intelligent feedback could help manufacturing enterprises proactively readjust the release of work orders that will be affected by uncertainties in order to improve the reliability and delivery of work orders.     

Latent variable iterative learning model predictive control for multivariable control of batch processes

A latent variable iterative learning model predictive control (LV-ILMPC) method is presented for trajectory tracking in batch processes. Different from the iterative learning model predictive control (ILMPC) model built from the original variable space, LV-ILMPC develops a latent variable model based on dynamic partial least squares (DyPLS) to capture the dominant features of each batch. In each latent variable space, we use a state–space model to describe the dynamic characteristics of the internal model, and an LV-ILMPC controller is designed. Each LV-ILMPC controller tracks the set points of the current batch projection in the corresponding latent variable space, and the optimal control law is determined and the persistent process disturbances is rejected along both time and batch horizons. The proposed LV-ILMPC formulation is based on general LV-MPC and incorporates an iterative learning function into LV-MPC. In addition, the real physical input that drives the process can be reconstructed from the latent variable space. Therefore, this algorithm is particularly suitable for multiple-input, multiple-output (MIMO) systems with strong coupling and serious collinearity. Three studies are used to illustrate the effectiveness of the proposed LV-ILMPC .     

Effect of upstream re-sequencing in controlling cycle time performance of batch processors

This paper demonstrates the effect of incorporating upstream sequence decisions in the cycle time control of batch processors. The focus is on dynamic control of a batch processor where near-future arrival information is available by predicting the process times on its upstream processor. In a particular decision point, the dynamic control task is to decide whether to start the batch process with one of the available product types immediately or to wait for future arrivals coming from the upstream processor. While considering a wait decision alternative, a re-sequencing approach at the upstream processor is proposed to shorten the arrival time of the desired product. Combining this architecture with a rolling horizon decision making, we propose a new look-ahead batch control strategy. A serial-batch processor system that is typical in wafer fabrication is modeled to demonstrate the approach. Simulation based benchmarking shows promising results.