Research On The Linear Hybrid Model Of The Operating Revenue Of Yunnan Country-level Electric Power Companies

As an important basic industry for the development of national economy,the development of electric power reflects the situation of economic operation to some extent.So it is necessary to analyze the operating income of electric power companies.Due to the actual access to electric power company is the longitudinal data,the traditional linear regression analysis and variance analysis of hypothesis conditions cannot be met and it is not suitable for the analysis of such data.Therefore,it is considered that the linear mixed model which is established in the form of random intercept and random slope is used to analyze the influencing factors of operating revenue of power companies.This paper starts with the principle of the model and introduces the form of the linear mixed effect model in detail.Then based on the Bayesian theory,the model is analyzed by the BRMS package in R.The model simulation analysis shows that the performance of the model with missing data is worse than that without missing data.Therefore,this paper discusses the problem of missing value processing and obtains the conclusion that the layered mean interpolation method is more effective in several common interpolation methods which lays the foundation for empirical analysis.Finally,in the empirical analysis,the data of 23 county-level power companies in Yunnan Province are used and the missing value of the data is first processed by layered mean interpolation.Then the linear mixed effect model is established to analyze the impact of the main business income factors.The analysis results show that in the main business income influencing factors,electricity sales and average unit price of electricity sales have a significant positive impact on the main business income.That provides a certain reference value for the overall development of power companies and improving economic benefits.