BP-based intelligent case retrieval method of high-rise structure form selection

Abstract

Establishing a case retrieval algorithm is the key step of the high-rise structure intelligent form optimization (IFO) based on case based reasoning (CBR). First, analyzes the defects of the traditional case retrieval method (CRM) based on typical distance and the characteristics of architecture structure case retrieval, an improved BP-based intelligent CRM was established and the topological structure and algorithm of the neural network was presented. Next, based on four target cases and twenty-five source cases, the training and forecast of the improved BP-based case retrieval was provided. Finally, the engineering application was given. The result shows: this intelligent CRM can overcome such shortcomings as poor learning ability, low searching efficiency and intelligent level, etc., providing comparatively more scientific method for the CRM of high-rise structure IFO