Application of the hottest wavelet method in trave

Application of wavelet method in traveling wave fault location of EHV transmission lines

Abstract: the traveling wave moving to the bus of substation will be generated after the fault of transmission line, so the traveling wave of fault current can be collected and recorded at the bus, and the accurate fault location of transmission line can be realized by using the fast algorithm of wavelet transform. However, due to the strong mutation information in the fault current signal of transmission line, it is necessary to use wavelet transform to detect the singularity of the real variable signal, so as to convert the time when the singular signal occurs into the fault distance. Through EMTP simulation calculation and detailed analysis of the results, this paper proposes a new method to calculate the fault distance by using the propagation of wavelet transform modulus maxima. Simulation results show that this method has high ranging accuracy

key words: wavelet transform transmission line singularity fault location power system relay protection 1 Introduction

there are two main types of fault location methods for EHV transmission lines at present [1,2]: impedance method and traveling wave method

impedance algorithm is based on the power frequency electrical quantity. It is a distance measurement method that calculates the line reactance between the fault point and the installation place of the distance measurement device by solving the voltage balance equation expressed in differential or differential form, and then converts the fault distance. According to the electrical quantity used, the impedance algorithm can be divided into single terminal electrical quantity algorithm and double terminal electrical quantity algorithm. No matter which algorithm is used, the impedance algorithm often cannot meet the accuracy requirements of fault location because of the influence of the error of the protective transformer and the transition impedance

the basis of traveling wave ranging method is that traveling waves have a fixed propagation speed (close to the speed of light) on the transmission line. According to this feature, measuring and recording the time when the traveling wave generated by the fault point reaches the bus when the line fails can realize accurate fault location. The early traveling wave method used voltage traveling waves, and theory and practice have proved that ordinary capacitive voltage transformers cannot convert traveling wave signals with frequencies up to hundreds of kHz. In order to obtain voltage traveling waves, special traveling wave coupling equipment needs to be installed, which makes the device complex and expensive, and lacks the technical conditions for measuring and recording traveling wave signals, and there is no appropriate mathematical method to analyze traveling wave signals, Therefore, it restricts the research and development of traveling wave ranging

wavelet analysis [3] as a safety protection measure of Mathematics: branch, with its theoretical perfection and extensive application, has attracted the attention of the scientific and engineering circles. At present, wavelet moisture analysis is gradually applied to power system because it will weaken the wear mark effect of polishing. Wavelet transform can be used to decompose the current and voltage signals with singularity and instantaneity obtained from fault recording, reflect the fault signal on different scales, determine the appropriate distance function according to the characteristics of the fault signal, and then solve the fault time and place that cause the sudden change of this signal, so as to realize fault location

2 mathematical model of power line

strictly speaking, the parameters of power line are uniform. At last, aluminum cables are basically distributed. Even a very short section of line has corresponding resistance, reactance, susceptance and conductivity (as shown in Figure 1). In general, what needs to be analyzed is usually the terminal condition, the voltage, current and power of the two terminals. Generally, the distributed parameter characteristics of lines can be ignored, and only in special cases can hyperbolic functions be used to study lines with uniformly distributed parameters. Although the existing distance measurement algorithm line models are diverse and have their own characteristics, in the final analysis, they all belong to the two line models of the enhancement number or distribution parameter of the ability of centralized participation to deal with trade protectionism

3 detection of singular signals by wavelet transform

if the function (signal) f (T) is discontinuous at a local point t0 or a derivative of a certain order is discontinuous, it is usually said that the function has singularity at t0 [3]. A more detailed description can be described by lipschtiz index:

let n be a non negative integer, and α Meet n ≤ α ≤ n+1, function f (T): [a, b] → R is Lipschitz at point x0 ∈ [a, b] α。 If there are normal numbers a and H0 and polynomial PN (x) of degree n, so that for any h ∈ (-h0, H0), there is

if there is α， F is not Lipschitz in x0 ∈ [a, b] α， Then the function f (T): [a, b] → R is singular at point x0 ∈ [a, b]

the definition of signal singularity is as follows:

let the function f (T): [a, b] → R, x0 ∈ [a, b], let α 0=sup{ α, F is Lipschitz at X0 α}， Then the Lipschitz singularity of F at x0 is called α 0。

obviously α= 1, the function (signal) is continuously differentiable; When 0 α 1, the smoothness of the function decreases when α= Only continuous at 0. α The smaller the value, the higher the singularity of F (T) at t0. This kind of function (signal) often appears in the signal research of power system, and the singularity of signal is often used to determine the time and cause of fault. The fast algorithm of wavelet transform can effectively propose the singularity of fault traveling wave

in Figure 1, in case of phase a ground fault, the current sampling values of phase a at the relay protection installation are

the sampling values of phase a current are obtained by wavelet transform, and two groups of values, smooth signal and detail signal, are obtained. Given the 2-scale sequences {pk} and {qk} of the orthogonal wavelet and its scale function, let the sampling sequence be {cj+1, k}, then the decomposed sequence is, and the transformation formula is

and its significance can be shown in Figure 2. That is, the sampled value is decomposed into smooth information C and detail information D by wavelet. Yes, the smooth letter obtained