论文部分内容阅读
结合文化算法的双层结构和多智能体进化算法的演化优势,提出一种求解资源受限项目调度问题的多智能体文化演化算法。算法设置了上层信仰空间和下层群体空间,各空间内智能体通过与其邻域进行竞争、合作操作及自学习操作来增加自身的能量,空间之间的交互是定期通过接受操作和影响操作采用同步传输方式来完成。通过对资源受限项目调度标准数据库PSPL IB中多个32、62、92、122工作的项目调度问题的仿真,结果表明:此算法不仅具有很好的收敛特性,而且运行速度快,是一种求解大规模调度问题的有效算法。
Combining with the double structure of cultural algorithm and the evolutionary advantage of multi-agent evolutionary algorithm, a multi-agent culture evolution algorithm for solving resource-constrained project scheduling problem is proposed. The algorithm sets the upper belief space and the lower group space, and the agents in each space increase their energy by competing with their neighbors, cooperating operations and self-learning operations. The interaction between the spaces is implemented periodically by accepting operations and influencing operations Transmission method to complete. Through the simulation of the project scheduling problem of multiple 32,62,92,122 tasks in the resource limited project scheduling standard database PSPL IB, the results show that this algorithm not only has good convergence characteristics, but also runs fast An efficient algorithm for solving large scale scheduling problems.