Bandwidth Scheduling Of Networked Control System Is Optimized

With the development and cross-penetration of computer technology, network communications technology and control science, NCS (networked control system) comes into being and is used widely. The performance of NCS is not only dependent on the control algorithm, but also on the scheduling of all kinds of data. Under the bandwidth restriction condition, the scheduling algorithm and optimization problem of networked control system are studied in this thesis. The primary content is as follows:(1) The optimization problem of NCS with multi-restraint and multi-objective is studied. Firstly, we analyze the necessity of optimization for NCS, and analyze the optimization objective function as well as the constraints. Secondly, we establish the model of bandwidth restriction as the objective function and the model of bandwidth restriction as the constraint condition. The simulation based on RM scheduling strategy verifies the validity of the optimization models.(2) Based on the limited bandwidth resources, one scheduling strategy is presented for periodic data. By using the time-window concept, this algorithm schedules the periodic data, and identifies the sampling period and the initial sampling time of each control closed-loop. In this algorithm the sensors use time-driven mode, and controllers use event-driven mode. Compared to the case that sensors and controllers both use time-driven mode, this algorithm can save more network bandwidth.(3) By using the time-window concept, one scheduling algorithm is designed to schedule the periodic data, time-critical data and time-available data. The periodic data's sampling period and the initial sampling time are determined, the limited transmission time of time-critical data and time-available data are given, and schedulability proof is given. The algorithm satisfies timeliness and bandwidth demand of each kind of data, and distributes the network bandwidth. Finally, the simulation case is given.The simulation results indicated that the optimized model can satisfy the band-width resources limitation and obtain good control performance; the scheduling algorithm can schedule each kind of data effectively, dispose the band-width resources reasonably.