论文部分内容阅读
如何有效满足连续型消费且消费量呈周期性波动的客户需要,是生产供应商急需解决的现实问题。本研究对连续型客户日消费量的时间序列数据,首次利用谱分析方法进行消费量的波动特征分析,再根据客户不同波动指标经过分层聚类分析以及判别分析,构建了判别函数方程。基于WE销售公司在江苏省的用气客户的日消费量数据进行了分析验证,研究发现,客户消费的波动特征可以通过谱分析得出的13项指标进行刻画反映,并根据这些波动指标对广大客户进行分类,以便供应商针对不同波动类别的客户采取不同的调峰手段,为供应商优化调峰方案提供分析基础,特别是根据新用户未来用气波动需求进行判别归类,有利于生产供应商提前主动地采取有效措施解决新用户的消费波动问题。
How to effectively meet the continuous consumption and cyclical fluctuations in consumer demand for customer needs is a real problem to be solved by suppliers. In this study, we analyzed the fluctuation characteristics of consumption for the first time using the time series data of continuous customer daily consumption, and then constructed the discriminant function equation through the hierarchical clustering analysis and discriminant analysis according to the clients’ different volatility indicators. Based on the daily consumption data of gas sales customers in Jiangsu Province of WE sales company, the study finds that the fluctuation characteristics of customer consumption can be reflected by the 13 indicators obtained by spectral analysis, and according to these fluctuations indicators, Customers are classified so that suppliers can adopt different peak shaving measures for customers in different fluctuation categories to provide an analysis basis for suppliers to optimize peak shaving plans. In particular, the classification and classification of new users based on future fluctuations in gas demand are favorable to the production and supply Businessmen take the initiative to take effective measures to solve the problem of consumer fluctuations in new users.