A price forecast model of hydraulic cylinders through integration of case based reasoning and back-propagation neural networks

Various kinds of hydraulic cylinders are widely used and have large variations in price. It has great significance to establish a hydraulic cylinder price forecast model for the issues of making procurement plans and the economical evaluation of maintenance and remanufacturing. However, it is hard to establish a direct mathematical model because there are many factors influencing the hydraulic cylinder price. The objective of this paper is to develop a price forecast model for hydraulic cylinders through the application of case based reasoning (CBR) and back-propagation neural networks (BPNN). Seven critical factors related to the hydraulic cylinder price have been found out by means of the Delphi method. They are cylinder bore diameter, stroke, pressure rating, servo cylinder or not, installation type, supplier qualification, and Producer Price Index (PPI). The PPI introduced in the model can improve the model sensitivity to market change. Case study is used to prove that both the BPNN model and the integrated CBR and BPNN model are capable of forecasting hydraulic cylinder price. The forecast of the integrated CBR and BPNN model is more accuracy than that of BPNN model.