Power supply enterprise faces the business risk caused by the power customers who break their promise on the supply contract. In order to avoid customer credit risks, power supply enterprise should set up customer credit management system and synthetically appraisal customer's credit risks. Power customers ' credit risk evaluation system involves many factors, as a complicated systems engineering. There are so many factors which influence the degree of power customers' credit risk. Meanwhile there are relations between the factors because it's difficulty for us to set up an index system which only has independent indexes. In other words, eliminating an index at discretion may reduce the evaluation accuracy of power customers' credit risk. But too many indexes will induce the complex of the computation. Thinking about the upper problems, the principal component analysis is introduced to deal with the indexes. It can form a new series of index which are independent among each other and reduce the index number. Then, the wavelet neural network (WNN) was used to calculate the degree of the power customers' credit risk. At last, it is validated that the results by this method is feasible for evaluation .
[1]
Qinghua Zhang,et al.
Wavelet networks
,
1992,
IEEE Trans. Neural Networks.
[2]
R. Kamimura,et al.
Selective responses by entropy minimization
,
1995
.
[3]
Liangyue Cao,et al.
Predicting chaotic time series with wavelet networks
,
1995
.
[4]
LI Jian-lun.
Empirical Research on Credit Risk Management for Electric Power Clients
,
2005
.
[5]
Huang Yan.
Assessment of power customer credit risk based on hiberarchy analysis method
,
2005
.
[6]
Matthew Lundberg.
Wavelets: A Tutorial in Theory and Applications (Volume 2 in series ‘Wavelet Analysis and its Applications’) and Wavelets and Their Applications
,
1998
.