An Intelligent Scheduling Method Based on the Allocation of Core Numbers for Tasks

This paper proposes an intelligent scheduling method based on the allocation of core numbers for tasks. The algorithm assigns different CPUs to tasks based on the size of the core numbers, established a Markov decision process (MDP) model for task allocation problem, and adopted the principle of resource allocation to maximize the reward value by balancing the processor resources. The algorithm searches for the optimal task allocation path by iteratively calculating the rewards and updating the Q values, completing the corresponding training cycle. Finally, the algorithm optimizes the output scheduling results based on the reward values. Through experimental analysis, it is shown that this algorithm can ensure the allocation of tasks is balanced while each task is allocated more reasonable according to the core numbers, maximizing the utilization of CPUs and thereby improving the overall system performance.