Research On Concrete Crack Analysis And Forecast System Based On Neural Network

Concrete is the widest applied building material in modern civil engineering which has been deeply studied and developed. Along with the adopting of new technology and the development of concrete material, the concrete crack is receiving more and more recognition as one important problem of concrete durable research. How to summarize the practical experience, develop new researching theory and construction technology for solving the concrete crack problem has become a stern topic which the general engineers and technicians faces.In view of this situation, this paper proposed applying the gray system and the neural network in the concrete crack analysis and the forecast, which is hope to provided one effective research method for the solution of concrete crack question.The main research contents of this article are as follows:1) Based on the summarizing of the familiar factor of crack, this article makes use of the language of Visual Basic to design a cause analytical system, which makes it becomes easy for the staff member to collect effective data and carry on further crack analysis.2) Two sets of concrete crack width forecast procedure combined with Visual Basic Language and the Matlab Neural Network toolbox are designed in this article. Users can carry out forecast based on the actual engineering data by choosing suitable forecast procedure.3) Gray system theory is used for the analysis of conjunction in the long-term forecast of concrete crack width. Conjunction analysis indicts the importance relations of the influencing factors.4) Using engineering example to test the concrete crack analysis and forecast system, Result indicate the system is feasible.