地下洞室安全监控软件系统的智能化问题

ARTIFICIAL INTELLIGENT TECHNICAL PROBLEMS IN SAFETY MONITORING SYSTEMS OF UNDERGROUND CAVES

结合小浪底安全监控系统的研发,探讨地下洞室安全监控系统的智能化技术问题。该系统对综合分析推理子系统中的测量因素分析、物理成因分析和综合分析评判3个子模块程序采用人工智能技术开发。系统根据领域知识的具体特点,分别采用不同方法提取知识,建造了4个不同类型的知识库。系统采用“原型系统”和“累进”设计开发方式,注意规则库中的深层领域知识的建造。

The purpose of the paper presented is to approach artificial intelligence techniques in development of underground structure safety monitoring system, in which artificial intelligence techniques was divided into three key subroutines: shouting and analysis for causes from surveying process, for causes from physical reason and synthestical analysis and reasoning in the subsystem. According to their specific characteristics of the knowledge domain, different ways were invoked to search the domain knowledge and to establish the four types of knowledge databases. 'Prototype modeling system' and 'increment development procedures' were adopted, which are in common with intelligent and decision supporting type software system. Concentration places not only on the construction of common rule database but also on the establishment of deep domain knowledge database. The approaches proposed were valuable and other safety monitoring system and underground structures.