Study On The Market Forecasting And Delivery Optimization Strategy Of The Cigarette Brand

In order to grasp the cigarette brand demand quickly and discover consumer's purchase tendency early, the tobacco commercial companies start using cigarette market forecasting technology gradually. A long time, the study of cigarette market forecasting mainly stayed in the forecasting of annual cigarettes sales volume, month's cigarettes sales volume, level changes and cigarettes price. But, for the municipal tobacco commercial companies'demands, the each cigarette brand sales forecasting is focus. In addition, China's cigarette market economy is the market economy under the national macro-control. In order to maintain healthy and orderly development of national cigarette market, higher levels of tobacco companies carry out performance evaluation to the lowers every year, which is also important to the municipal tobacco commercial companies. Therefore, based on the cigarette brand sales forecasting and the performance indicators, the making of brand delivery optimization strategy is another issue which these companies urgently need to solve. Combined with a municipal tobacco commercial company's actual demands, the study of these two aspects is conducted in this thesis.On the one hand, it is the research of cigarette brand market forecasting, that is cigarette brand sales forecasting research. Firstly, because cigarette sales forecasting is a forecasting of time series, the relevant contents of time series prediction were introduced in this thesis. Secondly, because cigarette sales have nonlinear characteristics, and the traditional statistical forecasting methods have limitations and neural network can reflect nonlinear characteristics, the ICBP neural network was used in this thesis, which is more general than the CBP network. The study found that cigarette sales have the characteristics of trend changes and seasonal fluctuations. According to these characteristics, the long-term trend model L_ICBP with considering social factors and the cycle model P_ICBP with combining time dependence were established in this thesis. Then, using the multiplicative model, the cigarette brand sales forecasting model ICBP was got. Finally, for a cigarette brand, a example of practical applications was given, and the cigarette brand market forecasting model was researched in this thesis.On the other hand, it is the research of cigarette brand delivery optimization strategy. Firstly, based on the theoretical knowledge of optimization method, software MATLAB with solving nonlinear programming was introduced in this thesis. Then, according to the municipal tobacco commercial company's performance indicators and the requirement of profits and taxes, the cigarette brand delivery optimization model was established, and the process of solving the optimal value was illustrated in this thesis.