In this paper, we propose a method for the design of an operational architecture of a Networked Control System (NCS). We consider a control system whose main goal is to control the position of a cart moving along a rail. The implementation of the controller is done through a distributed architecture in which a shared network supports the transmission of the samples between the sensor and the controller. For efficiently handling network congestion, we propose to apply a selective sample drop policy according to a (m,k)-pattern in order to decrease the network bandwidth required by this application during network overload periods. The paper shows how to determine the values of the parameter k that preserve the stability of the system and then how to identify the value of m and the (m,k)-pattern in order to optimize the system performance. IntroductionIn this paper, we consider a multi-variable plant and its computer-based control system. The input delivered to the plant is always the last reference computed by the controller (thanks to a zero-order-hold actuator). The output of the controlled system (the plant) is sampled (period h) and each sample is transmitted to the controller through a network. This will introduce a delay between the sampling instant and the control law completion. If this delay is constant, the value of this delay can be easily taken into account for the design of the control law. Nevertheless, the constant-delay assumption cannot be guaranteed if the network is a resource shared by several applications especially due to network overload period.For handling the network overload while observing the quality of control (QoC), two strategies can be applied:-The first one is based on the augmentation of the sampling period in order to reduce the required bandwidth. However, changing the sampling period of a controlled system alters its dynamics and needs a complex analysis of the system and its controller. -The second approach is to selectively drop some instances of system outputs in order to resolve the network congestion.This last solution is discussed in this paper and we propose, for a given kind of controlled system, a technique for the definition of the sample drop policy.More precisely, the proposed solution is based on the (m,k)-firm model [1], [2]. In this model, a recurrent activity is said to be under the (m,k)-firm constraint if at least m among k consecutive instances have to be processed before their deadline; m and k are integers and 0 m k < ≤ . Figure 1 shows that the effect of this policy is that some measures on the system are not transmitted to the controller. Consequently, this effect can potentially modify the Quality of the Control.Notice that most control systems can tolerate some misses of samples while preserving an acceptable level of QoC. So, this comment justifies identifying regular pattern of possible sample drops that can be cyclically repeated. Behind this idea, we focus our study on a (m,k)-firm approach; the problem is therefore to specify the controller...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.