Application Of Artificial Neural Networks In Displacement Monitoring And Prediction Of High-piled Wharf

With the development of the water carriage, loading and unloading tonnage ofwharf increases by about 20% per year. Wharf aging and dilapidation are seriousproblems hampering our water carriage field in the future. In response to thedevelopment of China water carriage trade, reducing aging and dilapidation of wharfcan insure wharf safety. It is urgent to set up a suit of scientific and maneuverablewharf maintenance and monitoring system. Dilapidation of the high-piled wharf ismore serious than that of other types and is easy to displace towards sea. Thus, inorder to guarantee the safety running of the port and to gain the optimal economic andsocial benefit, it is high time to analyze and estimate the displacement phenomenon ofthe high-piled wharf and take effective measures to prevent and control. ANN can simulate many frames and functions of organism, and is applied in theengineering and other fields. The ANN possesses the merits of rapid speed ofcalculation and characteristics of accepting the errors. Common studying method is toset up an idiographic model and the model must be the combination of linearity andnon-linearity of importance, but sometimes the model does not possess a good effectbecause of complexity of factor. The ANN has advantage on comparative with othermodel, so the ANN can be applied on displacement of high-piled wharf safetymonitoring. This paper designs a kind of new BP network and analyzes the example of the14th berth high-piled wharf of the second harbor in Tianjin port. The result indicatesthat the method of this text has good precision, adaptability and common ability andcan be widely applied to the field of displacement monitoring of water carriageengineering construction.