Refine
Departments, institutes and facilities
- Fachbereich Informatik (32)
- Fachbereich Ingenieurwissenschaften und Kommunikation (16)
- Institute of Visual Computing (IVC) (16)
- Institut für Cyber Security & Privacy (ICSP) (10)
- Institut für Verbraucherinformatik (IVI) (6)
- Institut für Technik, Ressourcenschonung und Energieeffizienz (TREE) (3)
- Fachbereich Wirtschaftswissenschaften (2)
- Fachbereich Angewandte Naturwissenschaften (1)
- Sprachenzentrum (1)
Document Type
- Conference Object (71) (remove)
Year of publication
- 2013 (71) (remove)
Keywords
- Education (2)
- Three-dimensional displays (2)
- end user development (2)
- 3D user interface (1)
- ARRs (1)
- Accessibility (1)
- Appropriateness (1)
- Cloud (1)
- Cognitive processing (1)
- Component Models (1)
Updating a shared data structure in a parallel program is usually done with some sort of high-level synchronization operation to ensure correctness and consistency. However, underlying synchronization instructions in a processor architecture are costly and rather limited in their scalability on larger multi-core/multi-processors systems. In this paper, we examine work queue operations where such costly atomic update operations are replaced with non-atomic modifiers (simple read+write). In this approach, we trade the exact amount of work with atomic operations against doing more and redundant work but without atomic operations and without violating the correctness of the algorithm. We show results for the application of this idea to the concrete scenario of parallel Breadth First Search (BFS) algorithms for undirected graphs on two large NUMA shared memory system with up to 64 cores.
Switched power electronic subsystems are widely used in various applications. A fault in one of their components may have a significant effect on the system’s load or may even cause a damage. Therefore, it is important to detect and isolate faults and to report true faults to a supervisory system in order to avoid malfunction of or damage to a load. If, in a model-based approach to fault detection and isolation of hybrid systems, switching devices are considered as ideal switches then some equations must be reformulated whenever some devices have switched. In this paper, a fixed causality bond graph representation of hybrid system models is used, i.e., computational causalities assigned according to the Standard Causality Assignment Procedure (SCAP) are independent of system modes of operation. The latter are taken into account by transformer moduli mi(t) ∈ {0, 1} ∀t ≥ 0 in a unique set of equations of motion. In a case study, this approach is used for fault diagnosis in a three-phase full-wave rectifier. Residuals of Analytical Redundancy Relations (ARRs) holding for all modes of operations and serving as fault indicators are computed in an offline simulation as part of a DAE system by using a bond graph model of the faulty system instead of the real one and by coupling it to a bond graph of the healthy system by means of residual sinks.