Updating industrial facilities to increase the level of automation and digitalization to match Industry 4.0 paradigms has become essential for many companies. Following such a trend, this paper presents a real-time optimization algorithm that plays a central role in a larger project framework devoted to highly interconnecting different network components of an Italian chemical industrial site. The proposed methodology aims at best managing the production rates of various products to fulfill a sales plan organized to satisfy numerous client requests. The considered model takes into account both batch and continuous processes as well as salable and non-storable products. The algorithm structure relies on the use of a non-linear optimization scheme and on the concepts of batch scheduling. Different features of the proposed methodology have been tested on real plant data, showing how the predicted forecast always improved the initial operation plan by considering both aspects of feasibility and economic nature. The use of the proposed algorithm assures the basis for fully integrating the control systems and the selling department of the facility in a more interactive and responsive manner.
A comparative study of different models and identification techniques applied to the quantification of valve stiction in industrial control loops is presented in this paper, with the objective of taking into account for the presence of external disturbances. A Hammerstein system is used to model the controlled process (linear block) and the sticky valve (nonlinear block): five different candidates for the linear block and two different candidates for the nonlinear block are evaluated and compared. Two of the five linear models include a nonstationary disturbance term that is estimated along with the input-to-output model, and these extended models are meant to cope with situations in which significant nonzero mean disturbances affect the collected data. The comparison of the different models and identification methods is carried out thoroughly in three steps: simulation, application to pilot plant data and application to industrial loops. In the first two cases (simulation and pilot plant) the specific source of fault (stiction with/without external disturbances) is known and hence a validation of each candidate can be carried out more easily. Nonetheless, each fault case considered in the previous two steps has been found in the application to a large number of datasets collected from industrial loops, and hence the merits and limitations of each candidate have been confirmed. As a result of this study, extended models are proved to be effective when large, time varying disturbances affect the system, whereas conventional (stationary) noise models are more effective elsewhere
7 January 2020 in Industrial & Engineering Chemistry Research
The paper presents the realization of a control loop performance monitoring system operating in cloud as single unity for a global supervision of data coming from various industrial plants located in different areas. This is a desirable solution for many companies owing to costs of local installations (systems, human resources and their maintenance and upgrading) and is made possible by available Industry 4.0 technologies, even though some aspects are worth of deeper investigation depending on specific industrial realities. The monitoring system (PCU – Plant Check Up) is described in its basic components and its evolution towards the cloud-based configuration. The whole architecture, as well as the solutions adopted for data acquisition from the field, transmission to the cloud and web app features which allow a full remote monitoring, are illustrated. Technical details about the application on a pilotscale plant are given and results for significant cases and different types of loop and actuator status are presented.
Valve stiction is one of the most common causes of poor performance in control loops. This paper presents a procedure which allows stiction quantification. The technique permits one to estimate the unknown real stem position, and moreover, it does not need any process knowledge and requires only the data normally registered in industrial plants. It is pointed out that the real problem consists of the lack of knowledge about the true value of stiction. A general methodology is proposed to discard data for which quantification is very likely to give wrong indications and to restrict its application to appropriate cases. Simulations show that several sources of perturbations can be eliminated, thus improving the reliability of stiction evaluation. Results are confirmed by application to industrial data: a significant number of valves are analyzed for repeated acquisitions before and after plant shutdown. The proposed procedure seems to be a valid methodology to monitor valve stiction and to schedule and check valve maintenance.