Boucle infinie | EPFL Graph Search

Une boucle infinie est, en programmation informatique, une boucle dont la condition de sortie n'a pas été définie ou ne peut pas être satisfaite. En conséquence, la boucle ne peut se terminer qu'à l'interruption du programme qui l'utilise. Pourquoi est-ce un problème ? Il y a rarement un intérêt à programmer une boucle infinie. Une telle boucle ne permet pas de faire sortir un résultat, et accapare les ressources de l'ordinateur. Sur un système monotâche, une boucle infinie peut interdire à l'utilisateur toute autre action. Il faut alors interrompre de force l'exécution (on peut être obligé de couper l'alimentation de l'ordinateur). Les systèmes modernes sont généralement suffisamment multitâches pour laisser le contrôle à l'utilisateur, qui peut commander l'interruption du programme. Mais tant que ce n'est pas fait, l'ordinateur se retrouve beaucoup moins performant pour tout autre travail. Une boucle infinie est donc presque toujours considérée comme un bug. Une ex

A Friction Compensation Control for Power Steering

Robert Fuchs, Philippe Müllhaupt

The effects of friction are critical to the dynamics of electric power steering (EPS). On the one hand, friction contributes to the stability of the system and filters some of the disturbances (road vibration, etc.). On the other hand, it affects negatively the driving feel and refrains from accurate positioning of the steering wheel. Also, for steering manufacturers, friction hinders the development and tuning of the assistance strategy. Therefore, being able to control or eventually suppress friction in EPS is a real issue. In this paper, a control strategy for the active compensation of friction in a column-assist type electric power steering (C- EPS) is presented. The assist motor and the electronic control unit are used to cancel the friction effect in order to imitate the behavior of an ideal, frictionless system. The feasibility of this strategy is demonstrated on the power column (the upper part of the steering system) using the same information (signals) as that available in an actual product. The proposed control is based on a model of the power column including slip and load dependent friction forces. For this purpose, a detailed simulation model, developed and validated in a previous work, is reduced into a lower-order model to enable real time computation. The LuGre model is used to compute both the static and dynamic friction forces with continuous formulation. The control architecture is composed of two cascaded feedback loops. The internal loop estimates the internal friction of the system and compensates it through the motor input. The external loop contains a frictionless reference model used as a trajectory planner and a linear controller which attempts to minimize the error between the plant response and the reference response. The stability and the robustness of the control strategy are analyzed formally. Specifically, it is shown that the limit error between the plant and the reference responses can be made arbitrarily small with appropriate values of the gains. Experi- mental results demonstrate that the control strategy is successful in tracking the frictionless, reference trajectory and confirm the robustness against inaccurate friction parameters.

Ieee-Inst Electrical Electronics Engineers Inc2015

Bridging the gap between dataplanes and commodity operating systems

Georgios Prekas

The conventional wisdom is that aggressive networking requirements, such as high packet rates for small messages and microsecond-scale tail latency, are best addressed outside the kernel, in a user-level networking stack. In particular, dataplanes borrow design elements from network middleboxes to run tasks to completion in tight loops. In its basic form, the dataplane design leverages sweeping simplifications such as the elimination of any resource management and any task scheduling to improve throughput and lower latency. As a result, dataplanes perform best when the request rate is predictable (since there is no resource management) and the service time of each task has a low execution time and a low dispersion. On the other hand, they exhibit poor energy proportionality and workload consolidation, and suffer from head-of-line blocking. This thesis proposes the introduction of resource management to dataplanes. Current dataplanes decrease latency by constantly polling for incoming network packets. This approach trades energy usage for latency. We argue that it is possible to introduce a control plane, which manages the resources in the most optimal way in terms of power usage without affecting the performance of the dataplane. Additionally, this thesis proposes the introduction of scheduling to dataplanes. Current designs operate in a strict FIFO and run-to-completion manner. This method is effective only when the incoming request requires a minimal amount of processing in the order of a few microseconds. When the processing time of requests is (a) longer or (b) follows a distribution with higher dispersion, the transient load imbalances and head-of-line blocking deteriorate the performance of the dataplane. We claim that it is possible to introduce a scheduler to dataplanes, which routes requests to the appropriate core and effectively reduce the tail latency of the system while at the same time support a wider range of workloads.

EPFL2018

Bridging the gap between dataplanes and commodity operating systems

Georgios Prekas

EPFL2018

A Friction Compensation Control for Power Steering

Robert Fuchs, Philippe Müllhaupt

Ieee-Inst Electrical Electronics Engineers Inc2015