Primary cut-out is important switchgear in power system, carries the task of electric network protection and control, and it is sent out
Raw malfunction and misaction often cause accident or expansion accident, cause huge economic loss and social influence.To realize safety
Stable power supply, power network proposes higher requirement to the reliability of primary cut-out.
The operational reliability of primary cut-out depends on effective condition monitoring and fault diagnosis (anomalous discrimination), and its is accurate
Property and practicality directly affect judgement for breaker performance state, base can be laid for follow-up Strategies of Maintenance by accurately differentiating
Plinth, so as to scientifically formulate repair schedule, improves utilization rate of equipment and installations, reduces maintenance cost, real on the premise of reliability is ensured
Existing economical operation.At present, the status monitoring for primary cut-out relies primarily on increase external sensor, using external sense skill
Art understands situations such as circuit breaker internal contact, mechanism, insulation, and this kind of method needs additional measurement device, and cost is high, and operation is multiple
It is miscellaneous, and easily by electromagnetic interference, therefore in actual applications, actual promotional value is not high for measurement process.
In fact, the realization of functions of the equipments is all the essence for reflecting equipment performance for the purpose of electric energy is transmitted in power network
It is the transmission performance of electric signal, the exception of its any equipment performance can be reflected by the change of electric signal.Thus,
Propose a kind of property abnormality method of discrimination for the primary cut-out for only depending on electric measurement information.Due to breaker normal condition
Descend and be failure to actuate, only circuit or capital equipment be disturbed or failure in the case of act, so with its corresponding recording number
It is suitable according to analysis breaker performance.
Arc time is the important parameter for indicating breaker arc-rupturing capacity, and the key accurately measured the arc time is
Determine the starting the arc moment of electric arc.The research work that breaker switch off time is measured mostly is opened with measuring breaker both at home and abroad at present
The form of disconnected arc time occurs.Representational technology is as follows：
First, according to the measuring circuit principle on branch wire time and the same period, building circuit and measuring.But this side
Method is suitable only for provisional monitoring.Secondly, the arcing moment of utilization space magnetic field signal on-line determination drop-out current.But this side
Method needs additional measurement device, and cost is high, complex operation, and measurement process is easily by electromagnetic interference, therefore in actual applications
Actual promotional value is not high.
The present invention is extracted the arc time (starting the arc moment to the quenching moment) as sign breaker in recorder data
The characteristic quantity of energy, because the judgement at starting the arc moment lacks corresponding switching value (or actual switch amount misregister), it is impossible to straight
Connect and obtain, the present invention finds the starting the arc moment using the method for wavelet packet analysis, high fdrequency component with special frequency channel is as characterizing
The characteristic quantity at arc moment, asks this moment, so as to obtain the arc time.Then, by counting a large amount of breakers normal arc time
Sample, analyzes its normal distribution law, delimits rational normal arc time scope, as the foundation of breaker anomalous discrimination,
Complete whole breaker anomalous discrimination process.
The content of the invention
The present invention is in order to solve the above problems, it is proposed that a kind of breaker based on wavelet package transforms and normal distribution law
Anomalous discrimination method, the present invention analyzes the fault characteristic value obtained, statistical analysis breaker arc time according to wavelet package transforms
Data, then select normal distribution to be fitted it, rational breaker arc time model divided according to the probability distribution of fitting
Enclose, be used as the distinguishing rule whether breaker is abnormal.
To achieve these goals, the present invention is adopted the following technical scheme that：
A kind of breaker anomalous discrimination method based on wavelet package transforms and normal distribution law, according to fault recorder data
The fault characteristic value obtained is analyzed, the data of statistical analysis breaker arc time are intended it using normal distyribution function
Close, rational breaker arc time scope is divided according to the probability distribution that fitting is obtained, sentence as whether breaker is abnormal
Further, the position of singular point is judged in the maximum of characteristic spectra mould according to wavelet package transforms, this is strange
Dissimilarity is exactly the starting the arc moment.
Further, when electric current vanishing, arc extinction obtains the quenching moment accordingly, and the starting the arc is subtracted with the quenching moment
At the moment, be the arc time.
Further, probability simulation is carried out to data statistics result with normal distribution, according to 3 σ principles, obtains normal arcing
Time range is within normal boundary value distance [μ -3 σ, μ+3 σ].
Further, the breaker arc time to be discriminated is obtained, then divides rational disconnected according to the probability distribution of fitting
Road device arc time scope, sees that the arc time whether in the range of the breaker reasonable arc time, judges whether breaker occurs
It is abnormal.
Compared with prior art, beneficial effects of the present invention are：
Present invention uses a kind of new method for being used to detect the breaker arc time, this method to recorder data by entering
The processing of row wavelet package transforms, obtains the arc time of breaker, real-time is good, and need not configure other devices.
The present invention determines a normal arc time model by the way that obtained a large amount of historical datas are fitted simultaneously
Enclose, and whether failure differentiates to breaker, it is clear and intuitive, for breaker fault distinguishing provide a kind of new method with
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
It is noted that described further below is all exemplary, it is intended to provide further instruction to the application.Unless another
Indicate, all technologies used herein and scientific terminology are with usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singulative
It is also intended to include plural form, additionally, it should be understood that, when in this manual using term "comprising" and/or " bag
Include " when, it indicates existing characteristics, step, operation, device, component and/or combinations thereof.
As background technology is introduced, there is the anomalous discrimination of primary cut-out in the prior art and supervised with online state
Based on survey, this method can reduce numerous drawbacks that plan repair is brought, and reduce the blindness of breaker maintenance, but need
The deficiency of numerous measurement apparatus is equipped with, in order to solve technical problem as above, is become present applicant proposes one kind based on wavelet packet
Change the breaker anomalous discrimination method with normal distribution law.
In a kind of typical embodiment of the application, as shown in figure 1, the present invention propose it is a kind of based on wavelet package transforms with
The breaker anomalous discrimination method of normal distribution law：
Step one, fault recorder data is collected, and wavelet package transforms processing is carried out to fault recorder data, when obtaining arcing
Step 2, according to the data statistics result of resulting breaker arc time, selects suitable probability distribution pair
It is fitted, and rational breaker arc time scope is divided according to the probability distribution of fitting, is sentenced as whether breaker is abnormal
Step 3, inputs fault recorder data to be discriminated, and the breaker arc time is obtained according to step one, according to step
Whether two obtained distinguishing rules differentiate extremely to breaker.
The method of step one is that fault recorder data is handled with wavelet package transforms, according to wavelet package transforms in spy
Levy the maximum of frequency range mould to judge the position of singular point, obtain the starting the arc moment.As shown in Figure 2.
The essence of wavelet transformation is to be referred to as the function Ψ (t) of wavelet by one to be after displacement b, then in different scale
Under a inner product is done with signal x (t) to be analyzed
In formula, a, b and t are continuous variables.B is time shift, and a is scale factor.
Its equivalent frequency-domain expression is
In this formula, x (ω), Ψ (ω) are x (t), Ψ (t) Fourier transformation respectively.
In the present invention, for the convenience of signal analysis and processing, signal will be discretized processing.Therefore, it is necessary to right
Yardstick a and displacement b carries out sliding-model control.
Wavelet transform is defined as
Wavelet packet analysis for a maximum advantage of wavelet analysis, be exactly can the more fine of signal analysis,
HFS is further segmented, so as to obtain the frequency domain of more refinements.2 are carried out for 5kHz original signal to sample frequency
After layer WAVELET PACKET DECOMPOSITION, its decomposition chart is shown in that the corresponding frequency sub-band of each node coefficient is shown in Table 1 in Fig. 3, decomposition texture.
The wavelet packet decomposition node of table 1 and frequency sub-band mapping table
When the contact of breaker is initially separated, the beginning of the breaker starting the arc represent.Due to electric arc electrical conductivity and
Sectional area all increases with the progress of arcing, especially electrical conductivity, it will produces acutely increased phenomenon, so just result in
The reduction of electric arc resistance.Therefore, when the breaker starting the arc, transient current has a certain size distortion, at the same time, and some are high
Frequency component can also be produced therewith.According to inspection information, the characteristic spectra of fault electric arc electric current is 782-1562.5Hz.More than being based on
Feature, the high fdrequency component that the present invention is produced with wavelet package transforms to arcing is extracted, and obtains the starting the arc moment.Specific method is root
The position of singular point is judged in the maximum of characteristic spectra mould according to wavelet package transforms, when the position of this singular point is exactly the starting the arc
Carve.When electric current vanishing, arc extinction, accordingly we can obtain the quenching moment.The starting the arc moment is subtracted with the quenching moment, just
It is the arc time.
The method of step 2 is to carry out probability simulation to data statistics result with normal distribution.According to 3 σ of normal distribution
Principle, judges normal arcing scope.
If it is μ that stochastic variable X, which obeys a mathematic expectaion, standard variance is σ2Gaussian Profile, be designated as：
X~N (μ, σ2),
Then its probability density function is
In actual applications, if one group of data is similar to the probability of normal distribution, the probability distribution for having 99.7% is existed
Distance average has the scope within 3 standard deviations.This is referred to as 3 σ principles.As shown in Figure 4.
P (μ -2 σ≤σ of x≤μ+2)=0.9544
P (μ -3 σ≤σ of x≤μ+3)=0.9972
In the present invention, after a large amount of breaker arc times are obtained, it can be seen that the data of statistics are almost into normal state
The breaker arc time, is fitted, according to 3 σ principles, is more than when the arc time for obtaining breaker by distribution with normal distribution
The σ of normal boundary value μ+3, it is possible to assert that its performance is present abnormal.
The method of step 3 is, according to the breaker arc time to be discriminated measured, whether breaker to be sentenced extremely
When obtaining circuit breaker failure recorder data to be discriminated again, its arc time is obtained according to step one, according to
The reasonable breaker arc time scope that the probability distribution being fitted according to step 2 is divided, sees whether the arc time closes in breaker
Manage in the range of the arc time, judge whether breaker exception occurs.
The preferred embodiment of the application is the foregoing is only, the application is not limited to, for the skill of this area
For art personnel, the application can have various modifications and variations.It is all within spirit herein and principle, made any repair
Change, equivalent substitution, improvement etc., should be included within the protection domain of the application.
Although above-mentioned the embodiment of the present invention is described with reference to accompanying drawing, not to present invention protection model
The limitation enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not
Need to pay various modifications or deform still within protection scope of the present invention that creative work can make.