Time-Space Modeling of Waste Electrical and Electronic Equipment Return Prediction

Jun LV,Shi-chang DU
DOI: https://doi.org/10.19495/j.cnki.1007-5429.2017.06.013
2017-01-01
Abstract:It is very critical to precisely predict waste electrical and electronic equipment (WEEE)returns in order to make reasonable recycling plans and design recycling network,etc.However,it is difficult to predict the returns of recycling WEEE since it has two characteristics in time-domain and space-domain simultaneously.The characteristics in time-domain and spacedomain of recycling centers and returns were described using stochastic variables.The spatial weight matrix of the prediction model was established by using topological graph method,and the space-time autoregressive moving average(STARMA)prediction model of WEEE returns was developed.The calculation process of space-time autocorrelation and partial autocorrelation were given,Yule-Walker equation in space-time form was derived,and the maximum likelihood method was used to estimate the parameters of the prediction model.Through case study of J company in Shanghai,the characteristics in time-domain and space-domain were analyzed effectively and the prediction error is less than 5 %,and the returns are predicted precisely.The proposed prediction model provides a new analysis method and theory foundation for WEEE returns.
What problem does this paper attempt to address?