摘要: |
Bridge design and construction processes consist of making models of the structures and then bringing the models into reality. In Bridge Management Systems (BMSs) a reverse operation is necessary, i.e., creating computer models of already existing structures. The variety of bridge constructions, as well as complex and changeable environmental, operational and economical conditions require advanced tools making the systems intelligent by equipping them with the ability of learning, recognizing, concluding, and even choosing and achieving goals. The proposed concept of the expert functions supporting main decision processes in the BMSs can fulfill the above mentioned requirements. The expert functions are based on data collected in the system as well as on the knowledge represented in the computer system. A prototype creation of the expert functions based on the artificial neural networks and analytic functions is described and the Bridge Evaluation Expert Function, which is being applied in the Railway Bridge Management System in Poland, is presented as a practical example of the proposed solution. |