Design of Cloud Task Scheduling Simulation Software Based on Multi-Agent Deep Reinforcement Learning

In order to solve the problem that traditional cloud task scheduling strategies cannot effectively respond to the allocation and optimization of computing resources in a dynamically changing cloud environment, this paper designs a cloud task scheduling simulation software based on multi-agent deep reinforcement learning. The software makes full use of the autonomy and cooperation characteristics of multi-agent systems, and realizes dynamic and adaptive allocation of cloud computing resources through real-time interactive learning and environmental feedback. The software is implemented on the python platform, which can achieve more efficient adaptive strategies, adjust parameters, and reduce the average maximum completion time of tasks for resource allocation and task scheduling problems in the ever-changing cloud computing environment.