Improving Flexibility in a Real-Time Fieldbus Network

ETFA 2001. 8th International Conference on Emerging Technologies and Factory Automation. Proceedings (Cat. No.01TH8597), 2001

The requirement for flexible operation is becoming increasingly important in many computer controlled systems, in order to cope with dynamic operational changes in a resource efficient way. This requirement naturally extends to the low level of such systems where, typically, field equipment such as sensors, actuators and controllers, are interconnected by a fieldbus communication system. This paper presents the DISCO project (DIStributed embeddable systems for COntrol applications) that started in September 2000, in the University of Aveiro, aiming at exploring specification, architectural and management issues in flexible distributed computer control systems for embedded applications. The project addresses the issues of control requirements definition in a QoS perspective, flexible scheduling of both network traffic as well as nodes' computing load, and global system management techniques. The paper shows the results obtained so far as well as the status of the current work in the scope of the DISCO project. Fig. 5. Timely error handling in FTT-CAN.

The paper examines the applicability of the new specification language E-LOTOS in the description of real time applications. E-LOTOS is the new extended version of LOTOS, currently under consideration by the ISO/IEC committee. The paper presents the complete process of producing the formal specification of a real time scheduler from its informal description. During this process, a semi formal model of the scheduler is first built that helps identify the critical parts of the scheduler. This model is then used as the basis for the formal specification of the scheduler. Finally, the degree at which E-LOTOs has succeeded in describing the critical parts of the scheduler is examined. The chosen, real life paradigm is the Link Active Scheduler (LAS) of the Fieldbus Foundation (FF) Data Link Layer (DLL) protocol. Its formal specification has not been presented before

2013

Scheduling theory, as it applies to hard-real-time environment, has been widely studied in the last twenty years and it might be unclear to make it out within the plethora of results available. Our goal is first to collect in a single paper the results known for uniprocessor, non-idling, preemptive/non-preemptive, fixed/dynamic priority driven contexts, considering general task sets as a central figure for the description of possible processor loads. Second to establish new results when needed. In particular, optimality, feasibility conditions and worst-case response times are examined largely by utilizing the concepts of workload, processor demand and busy period. Some classic extensions such as jitter, resource sharing are also considered. Although this work is not oriented toward a formal comparison of these results, it appears that preemptive and non-preemptive scheduling are closely related and that the analysis of fixed versus dynamic scheduling might be unified according to the concept of higher priority busy period. In particular, we introduce the notion of deadline-d busy period for EDF schedules, that we conjecture to be an interesting parallel of the level-i busy period, a concept already used in the analysis of fixed priority driven scheduling.

EXTENDED ABSTRACT: An Introduction to Time/Utility (or Time/Value) Functions and Utility Accrual Real-Time Scheduling, 2019

Extended Abstract (with References) This Introduction describes the TUF/UA paradigm and its default system model in more detail than previously documented. The default system model is based on application experience and generality. Any instantiation of the paradigm is application-specific and normally has a system model different from (a subset of) the default one. This paradigm has the side effect of illuminating the greater generality and applicability of "real-time" than is conventionally perceived. ¶ A Time/Utility Function (TUF)-originally [Jensen 77] [Jensen+ 85], and still often by others, called Time/Value Function-expresses the timeliness (both urgency and worth) of completing an action (such as a computational task or a device physical motion) in an application-specific utility ratio scale [Prasad+ 03], as an application-specific function of when (usually* in physical time) that action's operation completes. TUFs and their utility scales and values are derived from system-and application-specific subject matter knowledge (e.g., Clark+ 99]). A framework for assigning TUF utility values has been proposed for a class of very limited cases [Burns+ 2000]; but most cases necessitate specialized tools and techniques (e.g., [Lee+ 07]) *. ¶ Utility functions are normally non-convex-concave or linear (cf. Figure 2 in [Jensen 04]). Constant ones can either represent priority or be one way to represent relative importance. A (non-constant) TUF may have a "critical time" (a deadline is a special case), after which its utility does not increase [Locke 86]. A utility function's range may include negative values (penalties). A TUF's shape may be dynamically adapted-e.g., at an action or application mode change (such as for ballistic missile flight phases) [Maynard+ 88], cf. Figures 8 and 9 in [Jensen 04]. ¶ Actions which are being scheduled collectively may be any mix of aperiodic, sporadic, and periodic, but periodic actions do not receive traditional preferential treatment. Collectively scheduled (perhaps a subset of) actions are made to have consonant scales. At each of one or more scheduling instants, the scheduling algorithm considers the TUFs of all ready actions (and other system model information about the current situation), and schedules a set (from one to all) of those actions for operation. Specific UA scheduling algorithms are devised to provide specific acceptable timeliness and predictability-real-time QoS-assurances for specific system models and application situations. ¶ The scheduling is primarily according to two application-specific real-time QoS criteria-Utility Accrual (UA), which is the (commonly, expected) polynomial sum of their collective and individual utilities; and the (usually non-deterministic) predictability of that sum (i.e., of system* timeliness). Thus, UA scheduling is generally not greedy (and intentionally not fair), so actions may be either preemptible or not. A schedule normally also considers actions' constraints such as precedence and resource dependencies [Clark 90] [Li+ 06], and properties such as energy [Wu 05] and relative importance. ¶ In this paradigm, importance is application-specific (e.g., in terms of track quality, or weapon spherical error probable, etc.) and dynamic *. It is distinct from, and may be orthogonal to, an action's timeliness (urgency)-both

Annual Reviews in Control, 2007

A major goal of a supply chain is to ''improve the flow of material between suppliers and customers at the highest speed''. This goal suggests making real time and global management decisions, which involves scheduling. In this paper, we briefly remind the evolution of scheduling activities in industrial environment, describe the current situation and emphasize the problems that call for solutions. #

Massey University, 1995

Im diesem Dokument werden Aspekte der formalen zeitlichen Planung bzw. des Scheduling für Bauprojekte anhand ausgewählter Literatur diskutiert. Auf allgemeine Aspekte des Scheduling soll dabei nicht eingegangen werden. Hierzu seien als Standard-Referenzen nur Brucker (2004) und Pinedo (1995) genannt. Zu allgemeinen Fragen des Projekt-Managements sei auf Kerzner (2003) verwiesen.

In this note we propose an iterative scheduling technique which consists of consecutive forward/backward scheduling passes aimed at reducing the project duration by smoothing out the project's resource profile. The idea of iterative scheduling is initiated by Li and Willis in their related paper. The only common point between their scheduling technique and the one proposed here is the iterative feature. The two techniques differ both in algorithmic aspects and in the way activities are selected for scheduling at a decision point. In the technique proposed here activities are evaluated by well-reputed dispatching rules and a conflict based decision-making process called Local Constraint Based Analysis (LCBA). The results on benchmark problems from the literature demonstrate that LCBA specifically exploits the flexible activity time windows provided by the iterative scheduling technique.

2007 5th IEEE International Conference on Industrial Informatics, 2007

In a Foundation Fieldbus-based process control system, control strategies are implemented in a distributed environment on devices connected to the fieldbus. Although the fieldbus standard specifies the procedure to implement a given control strategy, it is left to system engineers to figure out the optimal configuration of a strategy. In this paper, we present a practical software tool that searches for the optimal deployment of a control strategy, in terms of overall finish time of the macrocycle and the number of published message on the bus. Despite the theoretically high worst-case time complexity, our tool's running time is acceptable in practice because of the limited size of a fieldbus network in practice. In addition, we propose an effective heuristic to reduce the running time further. Experiments with real world data show that our program outperforms a commercial control system in terms of the quality of schedules generated.

2000

The aim of this paper is to address the problem of correctly dimensioning real-time embedded systems scheduled with Fixed Priority (FP) scheduling. It is well known that computers which control systems are greatly affected by delays and jitter occurring in the control loop. In the literature, a deadline reduction approach has been considered as one solution to reducing the jitter