Data Mining Model For Bridge Health Monitoring Data

In the modernization drive of our country, the bridge-building is essential. Because of bridges'huge investment and the long usage, therefore the safety of its usage is extremely important. For all the bridges built, with the passage of time, for various reasons will bridge safety has been reduced, and impact vehicles'operation in security. There are many factors to impact the bridge security, such as the original design did not meet the use requirements, the construction did not meet design requirements, bridges disease, aging materials, fatigue effects without timely conservation, increasing vehicle load or traffic increases, the bridge deck joints damaged, and so on. With the economy development, made increasing demands in traffic, and reflected in the area of transport, the traffic volume is growing and vehicles carrying capacity is increasing, the traffic capacity of the bridge, carrying capacity, and functional disfigurement exacerbated. In order to ensure the normal use of the bridge the bridge health monitoring of the situation will become very important.At present the main bridge monitoring methods are divided into manual work method and automatic detection method, and they generally follow high testing cost, low efficiency and low sensitivity. In this paper, combining with technologies of data mining and the situation of bridge health monitoring system, a solution to handle large data set that bridge monitoring system generates is brought forward by the function of data mining technology in order to improve the analysis ability of bridge health state and to a great extent insures safe usage of the bridge. This major study is the usage of data mining technology to analyze bridge monitoring data and deal with a large amount of data that bridge-monitoring system collects in long-term monitoring stage. Established several models:(1) Clustering model, mainly for bridges data anomalies monitoring and data reduction.(2) Associating model, mainly for rules finding the between bridge structure and environmental parameters.(3) Time series analysis model, mainly for bridge monitoring data trends observation, the comparing of parameters changes trend, and the forecast of parameters value, the relevant parameters of the bridge can be found through parameters value time series chart comparison.