International symposium on process systems engineering最新文献

The development of condition monitoring systems often follows a modular scheme where some systems are already embedded in certain equipment by their manufacturers, and some are distributed across various equipment and instruments. This work introduces a framework for guiding the modular development of monitoring systems and integrating them into a comprehensive model that can handle uncertainty of predictions from the constituent modules. Furthermore, this framework improves the robustness of the modular condition monitoring systems as it provides a methodology for maintaining quality assurance and preventing unnecessary shutdowns in the event of some modules going off-line due to condition-based maintenance interventions.

Active control strategies play a vital role in modern pharmaceutical manufacturing. Automation and digitalization are revolutionizing the pharmaceutical industry and are particularly important in the shift from batch operations to continuous operation. Active control strategies provide real-time corrective actions when departures from quality targets are detected or even predicted. Under the concept of Quality-by-Control (QbC), a three-level hierarchical control structure can be applied to achieve effective setpoint tracking and disturbance rejection in the tablet manufacturing process through the development and implementation of a moving horizon estimation-based nonlinear model predictive control (MHE-NMPC) framework. When MHE is coupled with NMPC, historical data in the past time window together with real-time data from the sensor network enable model parameter updating and control. The adaptive model in the NMPC strategy compensates for process uncertainties, further reducing plant-model mismatch effects. The frequency and constraints of parameter updating in the MHE window should be determined cautiously to maintain control robustness when sensor measurements are degraded or unavailable. The practical applicability of the proposed MHE-NMPC framework is demonstrated via using a commercial scale tablet press, Natoli NP-400, to control tablet properties, where the nonlinear mechanistic models used in the framework can predict the essential powder properties and provide physical interpretations.

Rapid detection and identification of process faults in industrial applications is crucial to sustain a safe and profitable operation. Today, the advances in sensor technologies have facilitated large amounts of chemical process data collection in real time which subsequently broadened the use of data-driven process monitoring techniques via machine learning and multivariate statistical analysis. One of the well-known machine learning techniques is Support Vector Machines (SVM) which allows the use of high dimensional feature sets for learning problems such as classification and regression. In this paper, we present the application of a novel nonlinear (kernel-dependent) SVM-based feature selection algorithm to process monitoring and fault detection of continuous processes. The developed methodology is derived from sensitivity analysis of the dual SVM objective and utilizes existing and novel greedy algorithms to rank features that also guides fault diagnosis. Specifically, we train fault-specific two-class SVM models to detect faulty operations, while using the feature selection algorithm to improve the accuracy of the fault detection models and perform fault diagnosis. We present results for the Tennessee Eastman process as a case study and compare our approach to existing approaches for fault detection, diagnosis and identification.

A continuous rotary tablet press is a multi-stage process with many punch stations running in parallel, in which each punch undergoes the following steps: die filling and metering, pre-compaction, main-compaction, tablet ejection, and tablet take-off from lower punch. Process uncertainties or disturbances within a punch station or among stations in the tablet press are a major source of variation in final product quality attributes, e.g., hardness, weight, etc., which in turn imposes challenges for the real-time release in pharmaceutical continuous manufacturing of solid dosage. In this study, the direct compression line at Purdue University was investigated and a Natoli BLP-16 tablet press was used to characterize powder compressibility, system dynamics and variation, as well as the interaction effects on process control development. The compressibility of tablets made from a blend of Acetaminophen (API), Avicel Microcrystalline Cellulose PH-200 (excipient), and SiO₂ (lubricant) was found to be largely independent of tableting speed. By contrast, filling depth or dosing level, turret speed, feed-frame speed, and compression force were interacting and significantly affected the die-filling process and the final product quality attributes. Thus, the design of the process control structure plays an important role in reducing process and product quality variations. A hierarchical three-level control design was proposed and evaluated, consisting of Level 0 Natoli built-in control, Level 1 decoupled Proportional Integral Derivative (PID) cascaded control loops for tablet weight and production rate control, and Level 2 advanced model predictive control. Process variations, e.g., in powder bulk density changes, during continuous steady-state operation were also investigated. Finally, a risk analysis of the effects of the process dynamics on variation on the product quality control was briefly discussed and summarized.

The progress in the mechanistic understanding of the unit operations and the availability of multiple sensor technologies enable the inline implementation of data reconciliation and gross error detection methods in continuous pharmaceutical manufacturing. In this work, we demonstrate the benefits of accurate real-time monitoring of the process state in a continuous tableting process, with case studies representative of common situations in pilot-plant or manufacturing line implementation.

The land use allocation problem is an important issue for a sustainable development. Land use optimization can have a profound influence on the provisions of interconnected elements that strongly rely on the same land resources, such as food, energy, and water. However, a major challenge in land use optimization arises from the multiple stakeholders and their differing, and often conflicting, objectives. Industries, agricultural producers and developers are mainly concerned with profits and costs, while government agents are concerned with a host of economic, environmental and sustainability factors. In this work, we developed a hierarchical FEW-N approach to tackle the problem of land use optimization and facilitate decision making to decrease the competition for resources and significantly contribute to the sustainable development of the land. We formulate the problem as a Stackelberg duopoly game, a sequential game with two players - a leader and a follower (Stackelberg, 2011). The government agents are treated as the leader (with the objective to minimize the competition between the FEW-N), and the agricultural producers and land developers as the followers (with the objective to maximize their profit). This formulation results into a bi-level mixed-integer programming problem that is solved using a novel bi-level optimization algorithm through ARGONAUT. ARGONAUT is a hybrid optimization framework which is tailored to solve high- dimensional constrained grey-box optimization problems via connecting surrogate model identification and deterministic global optimization. Results show that our data-driven approach allows us to provide feasible solutions to complex bi-level problems, which are essentially very difficult to solve deterministically.