阅读说明：本技术 一种基于超平面聚类的pmu数据检测方法 (A kind of PMU data detection method based on hyperplane clustering ) 是由 薛安成 冷爽 严毅 徐飞阳 孔贺 于 2019-02-27 设计创作，主要内容包括：本专利提出了一种基于超平面聚类的PMU数据检测方法。该检测方法可以检测出在PMU数据有阶跃型偏差情况下的相角坏数据,从而辨别出可用的好数据。首先,获得线路在一段时间内的PMU测量数据,计算出正序分量；其次,使用超平面聚类对以线路传输有功功率为y轴,线路相角差为x轴的数据进行聚类；最后使用判据对聚类进行筛选,选取出正确数据。(This patent proposes a kind of PMU data detection method based on hyperplane clustering.The detection method can detecte out the phase angle bad data when PMU data has step change type deviation, to identify available good data.Firstly, obtaining the PMU measurement data of route whithin a period of time, positive-sequence component is calculated；Secondly, route phase angle difference is that the data of x-axis are clustered using hyperplane clustering to using line transmission active power as y-axis；Finally cluster is screened using criterion, selects correct data.)

1. a kind of PMU data detection method based on hyperplane clustering, feature comprise the steps of:

Step 1, installing PMU measuring device, acquire data sample;

Step 2 handles data, is clustered using hyperplane clustering to data;

Step 3 is judged using quality of the criterion to data.

2. the PMU data detection method according to requiring 1 based on hyperplane clustering, it is characterised in that: in the step 1, PMU device is installed at single back transmission line both ends, measures three-phase voltage, electric current phasor and the line at single back transmission line both ends The injection active power and reactive power at road both ends.

3. the PMU data detection method according to requiring 1 based on hyperplane clustering, it is characterised in that: in the step 2 In, wattful power is injected by three-phase voltage, electric current phasor and route both ends that PMU device measurement obtains single back line both ends Rate and reactive power；

Line taking road both end voltage phase angle differenceAs abscissa, route beginning injects active powerAs vertical seat Mark, make scatter plot (,);

DC power flow formula are as follows:

(1)

In ultra-high-tension power transmission line, X > > R, and can consider that route both end voltage is rated value, therefore formula can simplify are as follows:

(2)

Therefore can be classified at a distance from the random straight of mistake origin to data by calculating each point, i.e. hyperplane clustering; Wherein straight line are as follows:

(3)

Distance of the corresponding each point to straight line are as follows:

(4)

(wherein, k (j) indicates that jth class clusters center line slope)

Two classes are splitted data into, at this time center line by continuous iteration apart from size according to each cluster centre line of each pair of point It should be at the center of each cluster data.

4. the PMU data detection method according to requiring 1 based on hyperplane clustering, it is characterised in that: in the step 3, Judged according to quality of the criterion to data, that is, distinguish two classes cluster in which be available data;

Criterion: it is calculated using DC power flow formula (1), from the slope for obtaining cluster centre line in step 2(i= 1,2) index, is formed:

(5)

Formation criterion:

(6)

If the criterion of the i-th class dataIt is smaller, then it is believed that data category i preferably data.

Technical field

The invention belongs to Operation of Electric Systems to control protection field, and in particular to a kind of PMU number based on hyperplane clustering According to detection method.

Background technique

Currently, PMU data has become important one of the data source of smart grid.Carried out based on PMU/WAMS a series of Advanced application, such as parameter identification, state estimation, power system monitoring and control etc..The characteristics of PMU, is to be capable of providing markers one The Phasor Measurements data of cause, the accuracy of the data of PMU are the bases of all kinds of advanced applications.However, field measurement PMU data table There are quality problems for bright part PMU phase angle metric data；Meanwhile studies have shown that time synchronization, electric power mutual-inductor, PMU algorithm, The factors such as malicious attack may result in actual measurement PMU data, and there are deviations.It is existing studies have pointed out that severe day in terms of time synchronization Gas may interfere with GPS receiver device, and time synchronization is caused to generate deviation；GPS receiver loses the duration of satellite time transfer signal There are linear relationships between single-phase PMU phase angle measurements；GPS receiver can be interfered using countermeasure set, generate fixation and award When deviation.In systematic steady state, synchronism deviation Δ t will make PMU phase angle measurements in substation generate 2 π f₀The deviation of Δ t.This Outside, the ferromagnetic resonance of the capacitance type potential transformer under current intelligence, transient response problem, the saturation problem of current transformer, In PMU algorithm there is PMU data precision and influence in the factors such as spectral leakage, frequency fluctuation bring error problem etc..

For the credibility for guaranteeing the advanced application of PMU, accurate PMU metric data is crucial.Therefore, number is measured for PMU According to identification for use it is advanced application it is particularly significant.This patent is directed to the offset issue of PMU data, proposes a kind of based on super The PMU data detection method of plane cluster, it is intended to data deviation be distinguished and isolate correct data.

Summary of the invention

The PMU data detection method based on hyperplane clustering that the present invention provides a kind of, this method exist for PMU phase angle The case where deviation, will have deviation data and unbiased difference data to distinguish, to obtain precise information to be applied to electric system Control and protection.

Scheme is as follows:

A kind of PMU data detection method based on hyperplane clustering, feature comprise the steps of:

Step 1: installing PMU measuring device acquires data sample.

Step 2: data being handled, data are clustered using hyperplane clustering.

Step 3: improper data being judged using criterion.

In the step 1, PMU device is installed at single back transmission line both ends, measures the three-phase at single back transmission line both ends The injection active power and reactive power of voltage, electric current phasor and route both ends；

In the step 2, the three-phase voltage, electric current phasor and line at single back line both ends are obtained by PMU device measurement Active power and reactive power are injected in road both ends；

Line taking road both end voltage phase angle difference Δ θ=θ_m-θ_nAs abscissa, route beginning injects active-power P_mAs vertical Coordinate makes scatter plot (Δ θ (i), P_m(i))。

DC power flow formula are as follows:

In ultra-high-tension power transmission line, X > > R, and can consider that route both end voltage is rated value, therefore formula can simplify are as follows:

Therefore can be classified at a distance from the random straight for crossing origin to data by calculating each point, i.e., hyperplane is poly- Class.Wherein straight line are as follows:

P=k Δ θ (3)

Distance of the corresponding each point to straight line are as follows:

(wherein, k (j) indicates that jth class clusters center line slope)

Two classes are splitted data by continuous iteration apart from size according to each cluster centre line of each pair of point, at this time in Heart line should be at the center of each cluster data.

In the step 3, judged according to quality of the criterion to data, that is, distinguish two classes cluster in which be can Good data.

Criterion:

It is calculated using DC power flow formula (1)From slope k (the i) (i for obtaining cluster centre line in step 2 =1,2) index, is formed:

Formation criterion:

λ_ki=| 1- γ_ki| (6)

If the criterion λ of the i-th class data_kiIt is smaller, then it is believed that data category i is normal data.

Detailed description of the invention

In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this For the those of ordinary skill in field, without creative efforts, it can also be obtained according to these attached drawings other Attached drawing.

Fig. 1 is the PMU data detection method flow diagram provided in an embodiment of the present invention based on hyperplane clustering；

The simulated line topological structure that Fig. 2 is enumerated by the embodiment of the present invention.

Fig. 3 is the cluster result schematic diagram of the embodiment of the present invention.

Specific embodiment

Fig. 1 is flow chart of the invention, comprising the following steps:

Step 1: the data of route both ends PMU are obtained, positive-sequence component is extracted.

Step 2: to using line transmission active power as y-axis, phase angle difference at both ends of the line is that the track data of x-axis is surpassed Plane cluster, for the data deviation of step change type, can clearly obtain two categories.

Step 3: for both classifications, being differentiated using criterion, while meet two criterions is positive exact figures According to can be applied to the advanced application of PMU.

The method of the invention proved with example below.

The present embodiment builds 220kV analogue system using PSCAD, as shown in Fig. 2, to the positive sequence sequence of single time distribution line L1 Parameter is recognized.220kV route L1 is single loop line, line length 40.358km；Positive sequence parameter design value are as follows: resistance R₁= 1.625 Ω, reactance X_L1=12.289 Ω, susceptance B_L1=7.887e-5*2S.

Experiment one: (1) directly untreated using emulation data；(2) after using the plus noise on amplitude (I, U, Pm, Pn) Zero deflection emulate data, calculate its corresponding criterion, as a comparison.

Experiment two: after the deviation of phase angle, it is divided into two classes by hyperplane clustering, as shown in Figure 3.To these two types of data Criterion 1,2 is calculated separately, correct data is told.

By the criterion that provides it is found that second class cluster data is correct data in experiment two.With one correct data pair of experiment Than that can obtain, the criterion λ of secondary sources_k2Closer to experiment one data, test one criterion numerical value it is almost consistent, can sentence Determining secondary sources is correct data.

The method of the invention as shown in above-mentioned experimental result is suitable for the detection of PMU data, and this patent can be effectively The validity of PMU data is distinguished, and testing result has higher confidence level.

To sum up, the PMU data detection method of the present invention based on hyperplane clustering is feasible, and has engineer application valence Value.

It is worth noting that, the content being not described in detail in the embodiment of the present invention belongs to professional and technical personnel in the field's public affairs The prior art known.

The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto, Within the technical scope of the present disclosure, any changes or substitutions that can be easily thought of by anyone skilled in the art, It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claims Subject to enclosing.