Abstract:

 In the process of remanufacturing reverse logistics，there are many factors that affect the level of waste product recycling，the qualitative factors and quantitative factors，so the whole system is quite complicated. Based on statistical method of quantitative prediction modeling technology can't be used to forecast this kind of complex system incomplete，irregular data because this kind of model can't adapt to environmental changes or by the changing structure of the system itself caused by the nonlinear system. Therefore，it should be used to study the parameters of the non-network model to adjust the modeling technology to predict its uncertainty. In this paper, the fuzzy neural network theory is used to study the remanufacturfing reverse logistics model. Prediction models of waste product recovery time interval and the waste recovery yield are established. The two models can be adjusted by the data，which can directly obtain the change of the product sales，the amount of simulation and predict recovery yield with the change of time.

Key words: waste products, remanufacturing, reverse logistics, fuzzy neural network