| DC system plays a very important role in the safe and reliable operation of the power system.DC system of power plant and substation was important operating source of the power system.Dc operational power supply's backup power supply generally was battery.When the power of the system is cut off,the battery acts as backup power supply.As people's requirement to the power supply reliability and power quality is improved,and to ensure the safety and reliability of DC operational power supply,comprehensive on-line monitoring to battery operation parameters is necessary.Because there is closely correlation between residual capacity and resistance of battary,we can monitor the residual capacity of battary on-line comprehensivly by measuring residual capacity.At present though AC method can achieve on-line monitoring, this method has inherent defects.In order to eliminate the on-line monitoring noise, the wavelet is choosed to denoising in this paper which has the good localization characteristics in both time and frequency domains.Wavelet function is very important to wavelet analysis.Different wavelet function has different suitable scope and application field.At the same time, different scales of wavelet correspond to different frenquency ranges. In this paper,the select principle of wavelet fuction and scale are summarizeed,and B-spline function which is selected and Morlet fuction are introduced in detail.Based on deep research on the wavelet theory and the disadvantages of AC method to on-line monitoring,and according to the characteristics of interference noises, a new analysis method based sampling signal preprocessing by cubic B-spline wavelet and low frequency voltage value extraction by complex wavelet is put forward in this paper.This method can make full use of the characteristics of wavelet analysis,and remove the interference noises effectively.MATLAB was used to do the simulation in this paper.In the simulation, the new method that put forward above by choosing different Wavelet base functions and setting different parameters such as threshold value, scal and so on was analysed. The efferectness and accurate reliability of this method were proved by the result. Finally the design scheme of the on-line monitoring device based on DSP and CPLD was completed in this paper. |