摘要
将模式识别技术应用于桥梁结构状态评估与安全监测中,提出修正的欧氏距离与相似系数两种对响应信号特征向量进行统计分析的量化参数。提取出能较好反映出桥梁结构状态的特征参数,组合成特征向量。选择正常使用状态下多次实测数据特征向量的平均值作为基准状态,分别计算其他各次实测数据的特征向量,得到各次实测情况下特征向量与基准状态下特征向量之间修正的欧氏距离与相似系数。对桥梁正常运营状态进行监测,可得到修正的欧氏距离与相似系数的统计容许范围,作为桥梁结构状态评估的经验数字依据。
Mode recognition technique has been applied in the evaluation of bridge state and safety inspection.Quantizing parameter is presented between the revised Euclid distance and resemble coefficient respond signal feature vector. Characteristic parameter is extracted, which can better reflect bridge structure state, combining it into the feature vector. The average value of the normal state survey data feature vector is selected to be the reference state. Calculating other survey data feature vector separately, then can get the Euclid distance and the resemble coefficient between the normal state survey date feature vector and the reference state. Monitoring the motion of the bridge can get allowable range between the revised Euclid distance and the resemble coefficient which can be the experience digit basis for evaluating the state of bridge structure.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2005年第4期47-51,共5页
China Railway Science
关键词
模式识别
修正的欧氏距离
相似系数
桥梁结构
状态评估
量化评估
Mode recognition
Revised Euclid distance
Resemble coefficient
Bridge structure
State evaluation
Quantizing evaluation
