中文摘要：

软岩巷道锚注支护设计是一个涉及水文地质、工程地质、开采条件、岩石力学等诸多因素的复杂非线性难题。针对这一难题,将人工智能领域中专家系统技术与煤矿软岩巷道锚注支护领域专家理论研究成果、实践经验相结合,提出了软岩巷道锚注支护设计专家系统结构模型。基于软岩巷道锚注支护工程知识的特点,建立了软岩巷道围岩地质力学性质知识库、软岩巷道锚注支护设计工程案例知识库以及软岩巷道锚注支护设计专家知识库;同时,运用计算模式推理、BP神经网络推理以及产生式规则推理等3种推理策略,构建了锚注支护专家系统的核心推理机,实现了对软岩巷道锚注支护方案与参数的优化设计。将该专家系统应用于淮北矿区某矿井软岩巷道锚注支护设计中,显著地提高了软岩巷道锚注支护结构和参数选择的科学性与合理性,有效地维护了该软岩巷道围岩的稳定。

英文摘要：

The bolt-grouting support design for soft rock roadway is a complicated non-linear problem which relates to many factors, such as hydrogeology, engineering geology, mining condition, rock mechanics, etc. To solve this problem, expert system technology in artificial intelligence field has been combined with theoretical research results and practical experience of soft rock roadway bolt-grouting support, due to which an intelligent expert system model for bolt-grouting support design of soft rock roadways has then been proposed. Based on the knowledge of soft rock roadway support engineering, the database of geomechanical properties of soft rock roadway, the database of project cases of soft rock roadway bolt-grouting support design and the database of expert experience knowledge of soft rock roadway bolt-grouting support design have been established. Furthermore, by using three kinds of reasoning strategies（theoretical calculation, BP-ANN and production rule）, the core reasoning mechanism of the proposed expert system has been built to optimize the design of the parameters of bolt-grouting support for soft rock roadways. The proposed expert system has then been employed to design a soft rock roadway bolt-grouting support for a soft rock roadway in Huaibei coal mine area and the scientificity and reasonability of the selected parameters of bolt-grouting support for this roadway has been significantly improved. This system has also effectively maintained the stability of the soft rock roadway in the zone.