With the development of industrial internet, attention has been paid for edge computing due to the low latency. However, some problems remain about the task scheduling and resource management. In this paper, an edge computing supported industrial cloud system is investigated. According to the system, a constrained static scheduling strategy is proposed to over the deficiency of dynamic scheduling. The strategy is divided into the following steps. Firstly, the queue theory is introduced to calculate the expectations of task completion time. Thereupon, the task scheduling and resource management problems are formulated and turned into an integer non-linear programming (INLP) problem. Then, tasks that can be scheduled statically are selected based on the expectation of task completion and constrains of various aspects of task. Finally, a multi-elites-based co-evolutionary genetic algorithm (MEB-CGA) is proposed to solve the INLP problem. Simulation result shows that the MEB-CGA significantly outperforms the scheduling quality of greedy algorithm.
RGV Dynamic Scheduling Strategy Based on Network Cellular Automaton and Marko Model