Optimization of Preloaded Coefficient Based on the Adapted CBR

Selecting an appropriate preloaded coefficient has always been a challenge in wire- winding prestressed structure optimum design. Cased-based reasoning (CBR) has become a successful technique for knowledge-based systems in many domains. However, hardly any research has addressed the issue of how to generate the adaptation solution when the case has been retrieved. The present paper investigates the adoption of genetic algorithm(GA) to explore the suitable adjustment model. Two adapted model were presented and assessed in terms of their mean relative prediction error rates.The experiment results shown that applying GA to adjust the preloaded coefficient selection model is a feasible approach to largely improve the accuracy of estimation model. It also demonstrate that the adapted CBR presents better estimate accuracy than the results ontained by other unadapted CBR methods.