我国中小型大坝安全分析主要是对人工采集的监测数据进行后期离线分析,并结合丰富的实践经验和专业知识,判断大坝安全状态,该方法存在测量精度低、工作量大、实时性差等弊端。针对上述问题,提出了在基于B/S架构的大坝安全自动化监测系统上,结合Java Builder工具,调用Matlab算法,对大坝监测数据建立基于偏最小二乘回归算法的裂缝开合度统计模型,从而实现大坝安全实时分析。实验结果表明该系统能实现大坝实时、自动化安全分析,具有实用价值。
Doing some off-line post-analysis based on the collected monitoring data and depending on the rich experience and professional knowledge to judge the statement of dam are the method of analysis on small and medium dam in China,which has a disadvantage of low-accuracy,huge workload,bad real-time performance.In order to address the problem,a new heterozygosity of crack statistical model based on the partial least squares regression algorithm is proposed for the dam's monitored data,which is in the dam's safety system of automation monitoring based on the B/S structure and combining the Java Builder tools as well as calling the Matlab algorithms.The real-time analysis on the safety of the dam is realized by the model.The performance evaluation shows that the system can realize the real-time and automatical analysis for the dam's safety and has practical value.