Research on Improved Earliest Deadline First Scheduling Strategy of Networked Control System

Abstract

How to design an effective scheduling strategy to improve the performance of tasks in a networked control system, optimize the quality of data transmission and improve the adaptability of tasks with dynamic attributes is of great significance. Considering that the overall performance of networked control system is closely related to the time attribute of transmission task and channel environment, this paper improves the classic earliest deadline first strategy. Starting from the schedulable condition, the dynamic time interval of the task is precisely processed, and the distribution function of the schedulable condition is obtained by using the distribution function of each attribute, and the precise value of the schedulable variable that meets the requirements is determined, so as to obtain the accurate execution time and cycle value of the task attribute. The simulation experiment is carried out on the True Time platform. In the simulation, the scheduling effect of three loops networked control system is compared in ideal environment, with different packet loss rate and interference rate, and in the case of execution time and cycle change dynamically. The results show that the proposed improved earliest deadline first scheduling strategy overcomes the shortcomings of the standard earliest deadline first scheduling strategy, which leads to the degradation of system performance. It improves the output quality of the system, and improves the adaptability of the system to packet loss and interference factors.