Supervisory layer

The boundary of the safety-function should be limited to the first four layers of the data framework (plant, plant interface, reflex and supervisory layers [Faulkner 2002]) ... 

Process control in multienzyme processes, variables such as pH and temperature are often controlled during the process in order to reduce the influence that they produce on the dynamics of other variables and enzymes. For modeling, the controlled variables need to be identified in order to limit the capabilities of the model. In this case, they are included as assumptions of the model. Process control can be divided into two basic control layers [43]. The first is known as the regulatory layer, which controls variables such as pH and temperature. In this case, a simple controller design can be implemented. The second is known as the supervisory layer, which manages variables with more impact on the process such as concentrations of the compounds. In this case, a more detailed controller design is required. For multienzyme processes, this issue is highly relevant especially to achieve process improvements. 

Practical considerations in implementing the hierarchical control framework developed above concern the availability of manipulated inputs to address the control objectives in the slow time scale (it is possible that dim(us) < dim(ys)), as well as achieving a tighter coordination between the distributed and supervisory control layers. Both issues are effectively addressed by using a cascaded control configuration, which extends the choice of controlled variables in the slow time scale to include the setpoints y)p of the distributed controllers. 

Figure 3.4 Hierarchical control relies on separate, but coordinated, fast and slow controllers, designed on the basis of the respective reduced-order models, to compute the values of the separate inputs that influence the fast and slow dynamics of the process. Tighter coordination between the distributed and supervisory control layers is achieved by using a cascaded configuration.

The resulting hierarchical control structure is represented schematically in Figure 5.2. Note that, while controller design proceeds in a bottom-up manner, starting from the fastest time scale, during the operation of the process there will exist a tight top-down interconnection via control cascades between the supervisory and regulatory layers. 

A PAI, of course, is not statutorily required for approval of an application to market a drug product. A PAI may be requested by FDA headquarters s staff or conducted at the district office s discretion. Generally, if deficiencies are noted, at the conclusion of the PAI, the investigator will provide a letter to the company explaining its decision to recommend that approval be withheld until the center reviews the field s data. There are several layers of review within the agency after a PAI. The supervisory inspector, branch director, compliance officer, and district director may review any resulting deficiencies in the FDA district office. The FDA headquarters s review will be performed by the Office... 

Control System (BPCS), including functions of Supervisory Control and Data Acquisition (SCADA) system, the alarm system (AS) and Safety Instrumented Systems (SIS) performing defined Safety Instrumented Frmetions (SIF). Proper design of layers of protection is based on hazards analysis and risk assessment with consideration of human and organizational factors. It is essential to ensure required safety integrity level (SIL) for each of these layers. 

Obviously, from a purely mathematical point of view, it would be optimal to use a centralized on-line optimizing controller with continuous update of its model parameters and continuous reoptimization of aU variables. However, for a number of reasons, we almost always decompose the control system into several layers, which in a chemical plant typically include scheduling (weeks), site-wide optimization (day), local optimization (hour), supervisory/predictive control (minutes) and regulatory control (seconds). Therefore, we instead consider the implementation shown in Figure 1 with separate optimization and control layers. The two layers interact through the controlled variables c, whereby the optimizer computes their optimal setpoints (typically, updating them about every hour), and the control layer attempts to implement them in practice, i.e. to get c Cj. The main issue considered in this chapter is then What variables c should we control ... 

The design of the integration architecture used in the test platform described in Section 6 follows the CIM hierarchical structure that contemplates three main levels of aggregation planning and scheduling at the upper level, co-ordination at the intermediate level and supervisory control at the lowest level. This three layered architecture is represented in Fig. 10. 

In order to test and validate the developed architecture a case study has been designed and built. The test platform chosen allows validating the proposed methodologies with a simple case so that it ceui be transported to more complex production structures. The test problem includes the three decision-making layers planning and scheduling, co-ordination operation, supervisory control. 

Fig. 14. Bi-directional information flow across all layers (plant data, supervisory and coordination system, scheduling and planning).

Figure 7. Block diagram showing the interconnection supervisory control layer and the embedded control layer through a radio frequency link...

---