Research On Mobile Phone Recycling Network Based On Mobile Signaling Data Mining

In recent years, with the rapid development of economy and the improvement of living standards, the size of mobile phone users in China has reached 1 billion 530 million, but the annual recovery rate of less than 200 million mobile phones is less than 1%. The reason is that it does not form an effective recycling network of waste mobile phones for recycling and reuse of waste mobile phones. The most urgent problems to be solved in the research of the recycling network of waste mobile phone are: the prediction of the recycling of waste mobile phone; the location of the recovery point; the optimization of the recovery path. The mobile phone signaling data has the characteristics of high sampling rate, continuous data, low cost, so it can be used to carry out the research of the mobile phone recycling network. In this paper, the following four aspects of research work:(1) Analysis and construction of waste mobile phone recycling networkThis paper studies the recycling channels and methods of waste mobile phones abroad,and analyzes the current situation of recycling used mobile phones in china. It is concluded that the recycling of waste mobile phone network is characterized by uncertain demand,complex structure, diverse objectives and expensive disposal. The function of the recycling network is recycling, classification and detection, dismantling and reuse, waste disposal. The paper designs a recycling network of four layers of waste mobile phone, which is composed of recycling processing point, dismantling and recycling center, remanufacturing recycling and waste disposal center.(2) Prediction of spatio-temporal demand based on mobile signaling data miningStudy on the acquisition and format of mobile phone signaling data, analyzes the characteristics of mobile phone signaling data, design technical route recovery space demand forecasting, function factor function and space time factors are defined, and the use of mobile phone signaling data binding factors and spatial factors of two layer of the waste of time demand forecast for mobile phone itself, factors by using Bayesian inference, spatial and temporal factors by combining the stay point recognition, place of residence, work and activities to distinguish the spatial and temporal distribution curves depict the real data, finally use in Weifang city 5 days have been discarded mobile phone recycling demand in Weifang City, laid the foundation for the waste recycling mobile phone location.(3) Location model and solution of waste mobile phone recyclingThe construction waste recycling mobile phone location model, different is that the recovery needs to use with the traditional location model is the demand of mobile phone recycling waste of time and space, which is composed of two parts with time and space factors factors,takes the form of a temporal chain,that the number and location of demand are changing. The Lagrange relaxation algorithm and subgradient optimization algorithm are designed to solve the model. Finally the use of real data in Weifang for case analysis, according to the different candidate points,comparison of different location of the program number,it is proved that the model and algorithm construction has good practicability, can guide the enterprise location selection.(4) Waste mobile phone recycling path optimization model and solutionThe construction waste recycling path optimization model of mobile phone, and the difference lies in the traditional vehicle scheduling model: the number of recycling waste recycling point of mobile phone is uncertain; to solve the vehicle scheduling problem is a decision problem for a day,and the old mobile phone recycling processing path optimization problem is a long-term decision making problem. The hybrid Monte Carlo simulation and Hopfield neural network were used to screen the neural network, and the model was solved by genetic algorithm and neural network. Finally, the location of the candidate points 1200,the number of sites for the location of the 60 options for the typical case analysis. Through the convergence of the hybrid intelligent algorithm and the evolution of the recovery path, it is proved that the model and the solution algorithm can solve the practical problems,and guide the enterprise operation to improve the work efficiency.