CN104977506B - A kind of line fault recognition methods that classified analysis on major constituents is carried out using polar curve false voltage set of curves - Google Patents

Abstract

The present invention relates to a kind of line fault recognition methods that classified analysis on major constituents is carried out using polar curve false voltage set of curves, belong to Relay Protection Technology in Power System field.By closely to far along circuit setting monopolar grounding fault；External area error sets rectification side outlet failure, rectification side fault in ac transmission system, inversion side outlet failure and inverter side fault in ac transmission system.Under 10kHz sample rates, obtained respectively with line voltage set of curves under circuit external fault under the long scope internal fault in circuit all fronts by emulating, the data chosen in its line voltage set of curves 1ms carry out PCA cluster analyses, take two principal component PC₁And PC₂The 2 dimension PCA spaces formed.In this way, the two classes cluster point cluster of reflection line fault and external fault is formed on this PCA space.When breaking down, polar curve false voltage data in window during test data 1ms are mapped to thus 2 dimension PCA space, and the distance of test data and two class point clusters is measured using Euclidean distance, and form fault identification criterion.

Description

A kind of circuit event that classified analysis on major constituents is carried out using polar curve false voltage set of curves Hinder recognition methods

Technical field

The present invention relates to a kind of line fault identification that classified analysis on major constituents is carried out using polar curve false voltage set of curves Method, belongs to Relay Protection Technology in Power System field.

Background technology

The main so-called traveling wave based on change rate and variable quantity of the protection of direct current supply line that China has put into operation at present Protection, differential low-voltage protection, longitudinal differential protection and under-voltage protection etc..Wherein traveling-wave protection and differential low-voltage protection be Main protection, action protection speed is fast, but easily by interference effects such as harmonic wave, lightning stroke, sampled value shakes and is resistant to the energy of transition resistance Power is poor.It is the poor robustness because protection algorism easily by the reason for interference effects such as harmonic wave, lightning stroke, sampled value shake；With Voltage change ratio du/dt is that the traveling-wave protection of core criterion tends not to detection UHVDC circuits distal end high resistive fault, basic reason It is increase with transition resistance and the increase of fault distance so that du/dt reduces, in can not effectively be distinguished in time broadening Line-internal high resistance earthing fault and circuit external fault.And longitudinal differential protection and under-voltage protection are moved as back-up protection It is longer [4-8] to make the time limit.Longitudinal differential protection as high resistive fault detection is to escape the fault in ac transmission system phase by delay Between transient response to ensure its not malfunction, meanwhile, for not malfunction during avoiding power adjustment, in the design of SIMENS differential protections again Latching logic is added, so longitudinal differential protection outlet delay is too long, on the one hand, longitudinal differential protection may be caused usually Back-up protection effect is not had, on the other hand, DC control system protection may be caused first to act, cause unnecessary direct current Locking accident.

The content of the invention

During the technical problem to be solved in the present invention is external fault and line fault occurs for DC power transmission line circuit, It is proposed a kind of line fault recognition methods that principal component (PCA) cluster analysis is carried out using polar curve false voltage set of curves.

The technical scheme is that：

A kind of line fault recognition methods that principal component (PCA) cluster analysis is carried out using polar curve false voltage set of curves, By closely to far along circuit setting monopolar grounding fault；External area error set rectification side outlet failure, rectification side fault in ac transmission system, Inversion side outlet failure and inverter side fault in ac transmission system；Under 10kHz sample rates, circuit completely long model is obtained respectively by emulating Enclose under internal fault and line voltage set of curves, the data chosen in its line voltage set of curves 1ms carry out under circuit external fault PCA cluster analyses, take two principal component PC₁And PC₂The 2 dimension PCA spaces formed；In this way, reflection line is formed on this PCA space Road failure and two classes of external fault cluster point cluster；When breaking down, by polar curve false voltage data in window during test data 1ms Thus 2 dimension PCA space is mapped to, and the distance of test data and two class point clusters is measured using Euclidean distance, and forms failure Identical criterion.

Concretely comprise the following steps：

(1) sample database and its PCA Cluster spaces are established：By the failure completely set along UHVDC in long scope and set Rectification side failure and inverter side failure outside the area put, and obtained by electromagnetic transient simulation 1ms when window in polar curve false voltage it is bent Line cluster carries out PCA cluster analyses as sample data, PCA Cluster spaces is established, by PC₁And PC₂The PCA that reference axis is formed gathers Space-like forms the cluster point cluster of the obvious line fault mutually distinguished and external area error both modalities which；

(2) design and calculate the PCA cluster centres of sample data：The actual feelings of PCA Cluster spaces obtained according to step 1 Condition, clusters direct current transmission line fault and designs and choose k center (k >=1)；P is chosen to DC power transmission line external fault A center (generally taking p=1, also rectification side failure and inverter side failure can respectively be set 1 cluster centre)；

(3) after protection element fault initiating, polar curve false voltage data as test sample, are calculated and surveyed in window when taking 1ms The PCA Cluster spaces PC that sample notebook data is obtained in step 1₁、PC₂Projection o in reference axis_t(q₁, q₂)；

(4) projection o is calculated_t(q₁、q₂) with the Euclidean distance of each cluster centre：Calculate test sample data and line fault The Euclidean distance at k center of cluster, and d is denoted as respectively₁, d₂..., d_k；Similarly, calculate outside test sample data and circuit The Euclidean distance at p center of fault cluster, and d is denoted as respectively_k+1..., d_k+p；

(5) identification of line fault：Compare d₁, d₂..., d_k..., d_k+pSize, finds minimum value d_min；

If d_min=d₁(or d₂..., d_k), then it is judged as direct current transmission line fault mode；

If d_min=d_k+1(or d_k+2..., d_k+p), then it is judged as DC power transmission line external fault mode.

The beneficial effects of the invention are as follows：

(1) using PCA cluster analyses can effectively extract voltage curve cluster sample data main feature information and by its Principal component space is projected to, in PC₁And PC₂The different cluster point clusters of line fault and external fault are formed on coordinate space, are realized straight Flow Line failure and the effective of external fault portray, characterize and identify.

(2) based on test sample, Euclidean distance judges between the projection of PCA and the PCA clusters point all multicenters of cluster, is not required to Protection definite value is adjusted, criterion has adaptivity.

Brief description of the drawings

Fig. 1 is DC line analogue system.

Fig. 2 is the false voltage set of curves of cathode circuit measuring end：In figure, external area error has：Consider the event of rectification side outlet Barrier, transition resistance are set to 0 Ω, 10 Ω and 100 Ω；Rectification side fault in ac transmission system, including A phase earth faults, AB two-phases Earth fault, ABC three-phase ground failures；Inverter side failure includes inversion side outlet failure and inverter side fault in ac transmission system；A tables Timberline road false voltage curve, b represent external area error voltage curve.

Fig. 3 is cathode circuit measuring end false voltage set of curves and its cluster result on PCA space.

Embodiment

With reference to the accompanying drawings and detailed description, the invention will be further described.

Embodiment 1：Using the analogue system shown in Fig. 1, according to above-mentioned steps (1) and (2), line fault and outside are obtained Cluster result of the failure on PCA space is as shown in Figure 3.Choose cluster centre of 4 cluster centres as line fault, i.e., (- 1.7261, -0.3977), (- 1.2570, -0.2770), (- 0.9649, -0.2062), (- 0.8875, -0.1960) and 1 are outer Portion fault cluster center (- 0.2556, -0.2761).And the Euclidean at 4 centers for clustering test sample data and line fault Distance is expressed as d₁~d₄, the Euclidean distance table that test sample data cluster 1 center with external fault is d₅。

(1) d is calculated in the step in claims (3) and (4)₁₌0.0345、d₂₌0.5063、d₃₌ 0.8064、d₄₌0.8837 and d₅₌1.4921；

(2) step in claims (5) obtains d_min=d₁, so being judged as line fault.

Embodiment 2：Fault distance M end positions are 400km, 100 Ω of transition resistance.

(1) d is calculated in the step in claims (3)~(4)₁₌0.8024、d₂₌0.3182、d₃₌ 0.0199、d₄₌0.0602 and d₅₌0.6931；

(2) step in claims (5) obtains d_min=d₃, so being judged as line fault.

Embodiment 3：Rectification side outlet failure, 10 Ω of transition resistance.

(1) d is calculated in the step in claims (3)~(4)₁₌1.4920、d₂₌1.0082、d₃₌ 0.7082、d₄₌0.6302 and d₅₌0.1888；

(2) step in claims (5) obtains d_min=d₅, so being judged as external area error.

The embodiment of the present invention is explained in detail above in association with attached drawing, but the present invention is not limited to above-mentioned Embodiment, within the knowledge of a person skilled in the art, can also be before present inventive concept not be departed from Put that various changes can be made.

Claims (2)

1. a kind of line fault recognition methods that classified analysis on major constituents is carried out using polar curve false voltage set of curves, its feature are existed In：By closely to far along circuit setting monopolar grounding fault；External area error sets rectification side outlet failure, the event of rectification side AC system Barrier, inversion side outlet failure and inverter side fault in ac transmission system；Under 10kHz sample rates, circuit is obtained respectively completely by emulating With line voltage set of curves under circuit external fault under long scope internal fault, the data in its line voltage set of curves 1ms are chosen PCA cluster analyses are carried out, take two principal component PC₁And PC₂The 2 dimension PCA spaces formed；In this way, formed on this PCA space anti- Reflect the two classes cluster point cluster of line fault and external fault；When breaking down, by polar curve false voltage in window during test data 1ms Data are mapped to thus 2 dimension PCA space, and the distance of test data and two class point clusters is measured using Euclidean distance, and form Fault identification criterion.

2. the line fault according to claim 1 that classified analysis on major constituents is carried out using polar curve false voltage set of curves is known Other method, it is characterised in that concretely comprise the following steps：

(1) sample database and its PCA Cluster spaces are established：By the failure set along the long scope in UHVDC all fronts and setting Rectification side failure and inverter side failure outside area, and by electromagnetic transient simulation obtain 1ms when window in polar curve false voltage set of curves PCA cluster analyses are carried out as sample data, PCA Cluster spaces are established, by PC₁And PC₂The PCA clusters that reference axis is formed are empty Between form the cluster point cluster of the obvious line fault mutually distinguished and external area error both modalities which；

(2) design and calculate the PCA cluster centres of sample data：The PCA Cluster space actual conditions obtained according to step 1 are right Direct current transmission line fault cluster designs and chooses k center, k >=1；P center is chosen to DC power transmission line external fault；

(3) after protection element fault initiating, polar curve false voltage data as test sample, calculate test specimens in window when taking 1ms The PCA Cluster spaces PC that notebook data is obtained in step 1₁、PC₂Projection o in reference axis_t(q₁, q₂)；

(4) projection o is calculated_t(q₁、q₂) with the Euclidean distance of each cluster centre：Test sample data are calculated to cluster with line fault K center Euclidean distance, and be denoted as d respectively₁, d₂..., d_k；Similarly, test sample data and circuit external fault are calculated The Euclidean distance at p center of cluster, and d is denoted as respectively_k+1..., d_k+p；

(5) identification of line fault：Compare d₁, d₂..., d_k..., d_k+pSize, finds minimum value d_min；

If d_min=d₁, or d₂..., d_k, then it is judged as direct current transmission line fault mode；

If d_min=d_k+1, or d_k+2..., d_k+p, then it is judged as DC power transmission line external fault mode.