A Study on the PRISM-Based Scheduling Model for Multi-Constrained Systems

Abstract

In multi-task complex systems, how to achieve efficient task scheduling and optimal resource allocation under multiple conditions such as resource constraints, process constraints and time window limitations is a key issue that urgently needs to be solved in the fields of systems engineering and intelligent control. For typical scheduling scenarios with strong process dependence, multi-stage processes and heterogeneous resource structures, this paper proposes a priority-driven scheduling model (PRISM). Firstly, by constructing the temporal dependency graph model of the task process, the integer programming method is adopted to uniformly represent the task execution status and resource occupation. On this basis, a multi-dimensional priority evaluation function was innovatively designed. Combined with the dynamic resource state vector and the bidirectional scheduling control mechanism, the collaborative optimization of task execution periods and resource allocation was achieved. To deal with the problems of timing conflicts and resource competition in complex scenarios, an adaptive adjustment algorithm based on difference detection is proposed, which effectively guarantees the feasibility and robustness of the scheduling scheme. The experimental part was verified on a typical high-density task set. The results show that the proposed method can effectively improve the scheduling completion rate and resource utilization efficiency, and has good conflict repair ability and engineering adaptability. This method has the advantages of strong universality, stable scheduling performance and high deployability, and is suitable for the research of scheduling optimization problems in multi-type process systems.