论文部分内容阅读
原油储运中调度指令的生成过于依赖人工经验,导致作业方案难以最优,从而形成高库存、无效的资金使用以及运行费用的增加。同时,部分企业存在调度指令传达方式落伍,执行效率低、受众窄。本文基于扩展状态任务网,研究了调度指令优化算法并开发了相关软件,可以结合罐区现场实时运行数据,推理得到输送成本最小的指令集,并以网络化的方式下达。现场应用表明,所研发的系统具有很高的推理效率,并对罐区操作管理有促进作用。
The generation of dispatch instructions in crude oil storage and transportation relies too much on manual experience, which leads to the difficulty in optimizing the operation plan, resulting in high inventory, inefficient use of funds and increased operating costs. In the meantime, some enterprises have outdated modes of dispatching instructions, inefficient implementation and narrow audiences. Based on the extended state task network, this paper studies the scheduling instruction optimization algorithm and develops related software. It can combine with the real-time operation data in the tank farm area and deduce the instruction set with the lowest transmission cost and get it in a networked way. The field application shows that the developed system has high reasoning efficiency and can promote operation and management of tank farm.