Study On Satifaction Survey Grid Model And Trade Nethod

Reform of the power system in China gradually deepened,Electricity market has gradually become a buyer's market.how to do the work of customer satisfaction has become the focus of power supply enterprise.Customers for electricity and related services and products are satisfied has become one of the decisive factors in the development of enterprises.Power companies want to be able to really stimulate demand for electricity,effectively improve the management level,it should be the current level of service supply companies as a starting point,targeted improvements.This paper profound analysis of the macro model Customer Satisfaction Index and Customer Satisfaction Index Model ACSI,ECSI and ACSI,defines the relationship between the explanatory variables and latent variable models,as well as latent variables,constructed power customer satisfaction index model.Using structural equation modeling.Analyze the path coefficients.In the traditional system,the theoretical basis of satisfaction on the application of the second order closed-loop system and satisfaction with the theoretical combined system,interpretation of test coefficient system,which acts as an important power enterprise service level monitoring indicators.In this paper,multiple linear regression,principal component regression,a comparative analysis of the basic idea of the value of entropy weighting method,comparing the advantages and disadvantages among the principal component regression method and entropy weight Weighting combine application to power customer satisfaction index model,the regression coefficients to improve the system,but also reduces the problem of multicollinearity.Also studied standard management situation of China's power grid enterprises,from the power supply quality,standardized services,consulting services,ratepayers and service management five aspects to build benchmarking indicator system of new power grid enterprises with the industry;Again,the subject of research with the industry data mining algorithms,k-means clustering algorithm is very dependent initial value,the algorithm is easy to fall into local optima.K-means algorithm for the problems,we use PSO algorithm has strong global search and optimization features capacity is proposed based on k-means clustering algorithm improved particle swarm optimization algorithm.Firstly PSO global search,then k a global optimal cluster centers to its search as the initial cluster centers k-means algorithm,run k-means clustering algorithm to get the final results of divisions.verify the effectiveness and feasibility of the algorithm by the actual power of the company's standard;Finally,based on the same industry standard benchmarking results propose appropriate ratings and evaluation model for the grid decision support business development.