基于Multi-Agent技术的露天矿山生产调度系统框架的构建

针对于露天矿山生产调度系统的复杂非线性、工艺点多面广等特性,引进具有分布式管理特性的多智能体技术,建立基于多智能体技术的露天矿山生产调度系统.在系统中将生产调度分为任务Agent、生产调度Agent、爆破Agent、运输Agent、破碎Agent以及资源Agent六个单智能体.运用具有较强非线性拟合能力的自适应神经模糊推理系统(ANFIS)进行局部与整体的建模,并利用粒子群优化算法对模型进行优化,得出最优化生产调度方案.对矿山实际生产调度方案的制定具有一定的参考价值.

Framework of Open-pit Mine Production Scheduling System Based on Multi-Agent Technology

Considering the characteristics of complex nonlinear and processes over a broad area of the open-pit mine production scheduling system,the multi-agent technology with distributed management features is introduced to establish the production scheduling system of open-pit mine,which is based on multi-agent technology.The production scheduling system is composed of tasks agent,production scheduling agent,blasting agent,transportation agent,crushing agent and resource agent.The ANFIS with the stronger nonlinear curve fitting capability is used to model the local and global system.The model is optimized by particle swarm optimization algorithm,and then the optimal production scheduling program is obtained.The program provides certain reference value for the enactment of the actual mine production scheduling program.