【作者】 叶亚齐；

【导师】 许梦国；

【作者基本信息】 武汉科技大学 ， 安全技术及工程， 2004， 硕士

【摘要】 深基坑开挖与支护是一个非常复杂的系统工程，实践性较强，对实践经验的依赖也很高。工程技术人员在工程设计以及施工现场指导时往往很难把握，经常需要参考专家的指导、专业书籍的标准和以往成功的工程实例经验。为充分利用这些知识，本文力求在研究深基坑专业知识和专家经验知识的基础上，结合实际工程的应用，研制开发了理论可靠、实际可行的深基坑开挖与支护咨询专家系统。 基于知识的专家系统利用计算机模拟人类专家的思维方法，以有限的信息或在一定应用领域内正确做出与人类专家等同的判断和决策，给技术人员提供咨询指导。本文在研究与开发此系统过程中，主要进行了以下一些工作： ·根据本系统的应用领域和需要实现的功能，确定系统的结构、功能模块以及在开发过程中采取的策略。 ·通过领域专家、工程技术人员和相关专业书籍，总结深基坑开挖与支护所涉及的专业知识，用以构建专家系统的知识库。 ·分析专家经验知识的结构和专家系统表达此类知识的方法，选择适合本系统的知识表示方法，采用面向对象的方法表达并实现知识库的数据结构。 ·根据系统知识库的数据结构和知识的表达方式，采用广度优先搜索和深度优先搜索相结合的策略，建立正向推理机制以实现对系统知识库中专业知识的利用。 ·针对专家系统知识获取难的瓶颈，利用人工神经网络自学习的技术，实现本系统知识库的自学习功能，增强系统的知识获取的能力。 通过大量实例的运行以及与工程技术专家的比较，深基坑开挖与支护咨询专家系统完全达到了预期的目标，不仅提供的咨询结果正确，而且可以提供令人信服的解释。实践证明深基坑开挖与支护咨询专家系统的研究与开发具有很高的理论与实际应用价值。更多还原

【Abstract】 Excavating and retaining and protection of deep foundation excavations are highly complicated systems engineering, which need practicing very much and are depend on practice experience. Technical personnel are difficult to make decisions when they meet problems on the project design or on the spot of project constructing. So on such time, they need instruction of expert and specialty books and experience of anciently project examples. In order to make full use of above knowledge, this thesis develops the consultation expert system on excavating and retaining and protection of deep foundation excavations.On the base of knowledge, expert system simulates human expert’s thinking means with computer to solve problem, then supplies instruction to technical personnel. In the course of development, the thesis accomplish these works: According to the appliance field of the system and the function the system must have, the thesis fixes on the framework and function module and development policy of the expert system. In order to constructing the database of the system, this thesis summarizes the specialty knowledge on excavating and retaining and protection of deep foundation excavations with help of engineering experts and specialty books. Analyzing the configuration and the expression means of expert knowledge, this thesis chooses the knowledge expression means which is suit to the system, and constructs the data structure of the system with object orient method. According to the data structure of the system database, this thesis conjoins the breath-first search and the depth-first search, and constructs the positive direction reasoning mechanism to make use of the specialty knowledge in the database. Aim at the difficulty to add knowledge to the expert system database, the thesis realizes the self-learning function of the expert system with artificial neural networks.Validated by the running of a lot of examples, the consultation expert system on excavating and retaining and protection of deep foundation exactions has reached the development goal. The system can supply not only the right consultation result, but also the conviction explaining, which makes clear that the system has high value on theory and practice application.更多还原