初探大坝安全智能融合监控体系

System of theories and methods monitoring intelligently dam safety

大坝安全监控领域的多学科交叉的性质 ,要求将信息、人工智能、材料、工程与管理等多学科的理论和方法进行科学、有效、合理的融合 ,以期达到把目前广泛采用的离线、静态、被动的监控 ,转变为在线、动态、实时的智能监控目标。这是从根本上解决大坝结构整个寿命期间安全及减小灾害影响的一条崭新思路。本文从对大坝所具有的开放的复杂巨系统特性研究出发 ,探索性地研究了实现大坝安全智能监控目标的一条途径 -智能融合理论和方法体系。其核心思想是实现从单一正向思维模式向逆向思维、系统思维、反馈思维等全方位思维模式的转变 ,通过应用知识工程、信息智能处理等理论和方法 ,与传统的数学、力学方法相互渗透和融合 ,来完成大坝安全监控中的各项工作。作者初步将其分为数据级融合、分析级融合、诊断级融合和评价级融合四级体系结构 ,简要叙述了各级融合的基本原理和典型实现方法。

Offline and passive manner is conventionally adopted to monitor dam safety. From the viewpoint of system engineering, one dam can be regarded as an open, complicated and gigantic system. Its structural characteristics need be analyzed by the scientific integration of many subjects on information, artificial intelligence, material, engineering and management. A system implementing the fusion of theories and methods is presented in this paper. The system structure is composed of four fusion levels such as data, analysis, diagnosis and evaluation. The positive-thinking pattern is transformed into converse, systemic and feedback mode. The theories of knowledge engineering and information processing are crossed with traditionary mathematics and mechanics. The basic principles and representative methods on intelligent fusion are proposed in this paper. Dam safety can be monitored dynamically in real time by the proposed thought and methods.