Research On Reasoning Mechanism Of Intelligent Construction Control Expert System Of Cable-Stayed Bridges

Construction control is the key technology durning the constrution of cable-stayed bridge. Based on the previous researches of construetion control expert system, reasoning mechanism of the Intelligent Cable-Stayed Bridges’ Construction Control Expert System(ICBCCES) is researched. The research contents is as follows:1. The situation of Bridge Expert System and Reasoning Mechanism is analyzed in detail. The dissertation also summed up the current bridge expert system problems and the structure and organization of ICBCCES. At the same time, the significance of the ICBCCES reasoning mechanism for bridge problems and combination of cable-stayed bridge construction control feature and requirements has been explained.2. Reasoning knowledge base of ICBCCES is set up. The hierarchical semantic network representation and integrated knowledge representation method is adopted in this dissertation, and the qualitative knowledge in the process of cable-stayed bridge construction control is quantified.3. Reasoning mechanisms of ICBCCES is set up. The studies of reasoning mechanism can be divided into two parts, the first part was the reasoning ways of reasoning mechanism, which concludes the deductive reasoning, inductive reasoning,default reasoning and reasoning under uncertainty. The second is the strategies of reasoning mechanism, which included forward reasoning, backward reasoning, and hybrid reasoning. And some examples is used to illustrate the reasoning ways and strategies.4. The reasoning mechanisms of ICBCCES developed software. Microsoft Visual Studio2010 software is used as a development platform, Visual C++ as the development language and database access technology as Microsoft ADO.NET. It applied to the monitoring data Rugao Yangtze River Bridge construction control. The management of monitoring data, analysis and evaluation, control and decision-making exceptions warning are achieved in this system. The control processand the results show that it had reached the desired effect reasoning, and it wider range of applications.