卷烟品牌的市场预测和投放优化策略的研究

发布时间：2018-04-21 15:59

本文选题：卷烟品牌 + 市场预测　；参考：《东北大学》2009年硕士论文

【摘要】：为了迅速掌握卷烟市场的需求,尽早发现消费者的购买趋势,烟草商业公司逐渐地开始使用卷烟市场预测技术。长期以来,卷烟市场预测的研究主要停留在卷烟年销量预测、卷烟月销量预测、卷烟等级变化预测和卷烟价格预测这些方面。但是,对于市级烟草商业公司的工作需要来说,针对各个卷烟品牌的销量预测才是该公司预测工作的重点。此外,我国的卷烟市场经济是国家宏观调控下的市场经济,为了保持我国卷烟市场的健康有序的发展,每年上级烟草公司要对下级烟草公司进行业绩考核,这对于市级烟草商业公司也是非常重要的。因此,基于卷烟品牌销量预测和业绩考核指标的品牌投放优化策略的制定是该公司急需解决的另一个问题。结合某市级烟草商业公司的实际需要,本论文对这两方面进行了深入的研究。一方面,卷烟品牌市场预测的研究,即是卷烟品牌销量预测的研究。首先,因为卷烟销量预测是一种对时间序列的预测,本论文介绍了时间序列预测的相关内容。其次,因为卷烟销量具有非线性的特征,并且传统统计预测方法的局限性和神经网络具有能够反映非线性特征的特点,本论文采用比CBP神经网络更具一般性的ICBP神经网络。研究发现,卷烟销量具有趋势变化和季节波动的特点。根据这一特点,本论文建立了考虑社会因素的长期趋势模型L_ICBP,以及结合数据时间相关性的周期波动模型P_ICBP,从而使用乘法模型,便得到卷烟品牌销量的ICBP预测模型。最后,针对某卷烟品牌本论文给出了实际应用的例子,并且进行了卷烟品牌市场预测模型研究的探讨。另一方面,卷烟品牌投放优化策略的研究。首先,基于最优化方法的理论知识,介绍了具有求解非线性规划功能的MATLAB软件。然后,根据该市级烟草商业公司的业绩考核指标和利税要求,建立卷烟品牌的投放优化模型,并通过例证来说明求解最优值的过程。
[Abstract]:In order to quickly grasp the demand of the cigarette market and discover the purchasing trend of consumers as soon as possible, tobacco commercial companies gradually began to use the cigarette market prediction technology. For a long time, the research of cigarette market forecast mainly focuses on the forecast of annual sales volume of cigarettes, the forecast of monthly sales volume of cigarettes, the forecast of cigarette grade change and the forecast of cigarette price. However, for the job needs of municipal tobacco companies, the sales forecast for individual cigarette brands is the focus of the company's forecast work. In addition, the cigarette market economy of our country is the market economy under the state macro-control. In order to maintain the healthy and orderly development of the cigarette market in our country, every year, the superior tobacco companies should carry on the performance appraisal to the subordinate tobacco companies. This is also very important for municipal tobacco companies. Therefore, the formulation of brand placement optimization strategy based on cigarette brand sales forecast and performance evaluation index is another problem that the company urgently needs to solve. According to the actual needs of a city tobacco commercial company, this paper has carried on the thorough research to these two aspects. On the one hand, the research of cigarette brand market forecast is the research of cigarette brand sales forecast. Firstly, because the prediction of cigarette sales is a kind of time series prediction, this paper introduces the related content of time series prediction. Secondly, because cigarette sales has nonlinear characteristics, and because of the limitations of traditional statistical forecasting methods and neural networks can reflect the characteristics of nonlinear characteristics, this paper uses a more general CBP neural network than the ICBP neural network. The study found that cigarette sales have the characteristics of trend change and seasonal fluctuation. According to this characteristic, this paper establishes the long-term trend model LIC ICBP, which takes into account social factors, and the periodic fluctuation model, PICBPs, which combines the time dependence of data. Thus, by using the multiplication model, the ICBP prediction model of cigarette brand sales is obtained. Finally, a practical application example of a cigarette brand is given, and the market prediction model of cigarette brand is discussed. On the other hand, the study of cigarette brand placement optimization strategy. Firstly, based on the theoretical knowledge of optimization method, the MATLAB software with the function of solving nonlinear programming is introduced. Then, according to the performance evaluation index and the requirement of profit and tax of the city tobacco commercial company, the optimization model of cigarette brand is established, and the process of solving the optimal value is illustrated by an example.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2009
【分类号】：F224;F273.2;F426.8

【相似文献】