Study Of Coal Mine Distribution Network Voltage Sag Detection Method

The accurate detection and id entification of the voltage sag can help to ev aluate the regional distribution a nd mitigate the m easures,and can ef fectively analyze,com pensate and suppress the actual voltage sag,which is the m ain basis for the coordination betwee n the power sector and the power users.In order to sol ve the prob lem of feat ure extract ion of volt age sag signal and t o detect th e low accuracy of detecti on,the characteristics of voltage sag signal are expressed by wavele t entropy.According to the characteristics of fault wa velet entropy caused by different fa ult causes,Vector machine(SVM)classifie r to detect and i dentify.The wave form data of the voltage sag caused by various fa ults are simulated by Matlab / sim ulink.The data of the voltage sag signal is obtained for di fferent fault adjustment equipment parameters,and the wavelet entropy m easure is calcula ted.As t he classification of di fferent voltage sag f aults Standard,the final realization of the voltage sag of the autom atic detection of fast detection.The results show that the results are in co ntrast to the BP neural network m ethod,a nd the results show that there are obvious advant ages both in training time and in ide nti fying the accuracy rate.In a ddition,the im pact of underground mine noise is ve ry large,in the original signal by addi ng Gaussian white noise simulation of th e actual environm ent,the results show that anti-interference ability is very strong.The meth od of wavelet entropy combined with support vector machine can realize the fa st and accura te detection and recognition of voltage sag.