模式识别法在岩石边坡监测中的应用

提出并定义了边坡滑动模式的概念，将其与边坡安全监测资料分析结合起来，建立了监测信息与边坡滑动模式之间的映射关系。引入人工神经网络模式识别方法，建立了一个边坡实时安全监测模式识别器，根据监测资料实时判断边坡所处的安全状态模式，结合边坡的其他基本资料判断边坡的稳定性。作为实例，应用此模式识别器分析了小浪底水利枢纽出水口高边坡某一段时间的监测资料。实例表明该方法和工具的应用效果是值得信赖的，可满足边坡实时安全监测的需要。

APPLICATION OF PATTERN RECOGNITION METHOD TO SLOPE MONITORING

The concept of slide patterns is put forward. By use of slide patterns for slope safety monitoring, the mapping relations between slide patterns and monitoring information can be built. Through introduction of the pattern recognition method of artificial neural network, a pattern recognizer for slope monitoring is built to judge the real-time safety state of the slopes on the basis of slope monitoring information. The stability of slopes can be estimated based on the above information. As an example, this pattern recognizer is applied to the analysis of a section of monitoring data of Xiaolangdi outlet slope. It is shown that this pattern recognizer is reliable enough for slope monitoring.