摘要
利用计算机虚拟现实技术结合专家系统进行应急演练推理己成为化工安全演练方面的一个重要研究方向。然而传统的专家系统在面对新环境下的应急预案需要重新收集大量预案知识进行规则库的建立,缺少一定的适应性和流动性,无法满足化工企业根据新预案快速准确进行演练的要求。为了解决这一问题,本文提出了EFSM(Extended Finite State Machine)的组态化应急演练推理机制,主要设计思路为通过对基于文本的应急救援预案进行解析及可视化,利用组态化NPC(Non-Player Contro1)推理模块和基于EFSM的组态化真人推理模块分别负责处理应急演练过程中NPC的演练流程和真人角色的演练流程,形成完善的应急救援演练推理系统。最终将应急演练推理系统通过脚本的驱动,结合当前的虚拟现实三维技术进行应急救援演练仿真。
It has been an important study field in chemical industry to combine the virtual reality and expert system together to conduct the emergency drill. However, the traditional expert system needs to recollect the knowledge of the emergency plan when facing the new drill situation to build the rule base which lacks adaptability and flexibility. To solve this problem, this paper offers a new configurable emergency drill reference mechanism based on EFSM. The mechanism consists of three main modules. 1. Analysis the emergency rescue plan based on text. 2. Make use of the configurable NPC reference module to solve the NPC drill process. 3. Make use of the EFSM module to solve the real person's drill process. Finally, we combine the emergency system and virtual reality technology together to simulate the whole drill process.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2014年第7期838-842,共5页
Computers and Applied Chemistry
基金
北京市自然科学基金资助项目(4142039)
北京市优秀人才培养资助项目(2013D009016000004)
中央高校基本科研业务费(JD1310)
关键词
应急演练
EFSM
组态
推理
emergency drill
EFSM
configuration
inference engine