CN113484685A - Power grid fault diagnosis method based on time sequence organization type P system - Google Patents

Abstract

The invention discloses a power grid fault diagnosis method based on a time sequence organization type P system, which combines the time sequence characteristic of an alarm signal with the organization type P system, can better process the problems of false alarm, missing alarm and the like in the alarm signal, can also reason out the time interval of the fault occurrence moment and accurately describe the fault evolution process.

Description

Power grid fault diagnosis method based on time sequence organization type P system

Technical Field

The invention belongs to the technical field of power grid fault diagnosis, and particularly relates to a power grid fault diagnosis method based on a time sequence organization type P system.

Background

The power system fault diagnosis is intended to identify a faulty element and evaluate the action behavior of a protection device through an alarm signal obtained by a control center. The fault alarm information is an important precondition for accurate and effective fault diagnosis of the power system. The fault alarm signal of the power system usually retrieves an operation signal from a Supervisory Control And Data Acquisition (SCADA) system, but with the gradual expansion of the scale of the power system And the increasing complexity of the structure, a large number of alarm signals are rushed into a dispatching center, And when the alarm signals are too many, the operation such as misjudgment And missing judgment is easy to occur to a dispatcher. Therefore, there is an increasing demand for rapid and accurate fault diagnosis of the power system.

Up to now, various fault diagnosis methods have been proposed at home and abroad, for example: expert systems, Petri nets, artificial neural networks, and the like. From most of the existing fault diagnosis methods, the following problems are also faced: (1) most of the existing fault diagnosis methods still cannot well process the situations of false alarm, missed alarm and the like of alarm signals; (2) the fault diagnosis model based on the traditional organization type P system has low interpretability on the action sequence and logic between protection devices, and the model complexity still has a reduced space; (3) nowadays, with the gradual expansion of the scale of a power grid, a power system needs to cope with various fault scenes when a fault occurs, and how to efficiently and accurately identify a fault element under the various fault scenes becomes a key of a fault diagnosis method.

The accuracy and the rapidity of the fault diagnosis of the power system have important significance for rapidly recovering power supply and maintaining the safe and stable operation of the power grid system after the fault occurs. Therefore, a great deal of research needs to be done on how to solve the situations of false alarm, missing alarm, and the like of fault alarm signals when the power grid is in actual operation and describe the action sequence and logic among the protection devices when the power system is in fault diagnosis.

Disclosure of Invention

Aiming at the defects in the prior art, the power grid fault diagnosis method based on the time sequence organization type P system solves the problems of false alarm, missing alarm and low diagnosis efficiency of fault signals in the conventional power grid fault diagnosis.

In order to achieve the purpose of the invention, the invention adopts the technical scheme that: a power grid fault diagnosis method based on a time sequence organization type P system comprises the following steps:

s1, determining a fault area in the power grid to be diagnosed through a junction analysis method, and further determining a suspicious fault element;

s2, in SCADA-based systemConstructing and correcting initial configuration vector W for action information and time sequence information of protection device₀；

S3, establishing a TTPS fault diagnosis model of each suspicious fault element according to the action sequence among the protection devices;

s4, establishing a connection matrix C and an initial configuration vector W according to the TTPS fault diagnosis model₀Constructing a translation vector

S5, generating a fault node vector S according to the node distribution of each suspicious fault element in the TTPS fault diagnosis model;

s6, according to the transfer vector W₁And a fault node vector S, and constructing a judgment vector W_endAnd according to the judgment vector W_endThe non-zero elements in the system determine the elements which send out the faults, and the fault diagnosis of the power grid is realized.

Further, the step S1 is specifically:

s11, setting the initial iteration number k to be 1;

s12, numbering each element in the power grid to be diagnosed in sequence, and constructing an element number set P_k；

S13, from set P_kIn which an arbitrary component number is taken and placed in a subset Q of component numbers_kJudging whether a closed circuit breaker exists in the power grid to be diagnosed and is connected with the currently selected element;

if yes, go to step S14;

if not, go to step S15;

s14, finding out all elements connected with the breaker in the power grid to be diagnosed, and adding the serial number of the elements to the set Q_kAnd returns to step S13;

s15, increasing the value of the search iteration number k by 1;

s16, collecting Q_k-1From the set P_k-1Get a new component number set P_k；

S17, judging the current new component number setP_kWhether it is empty;

if yes, go to step S18;

if not, returning to the step S13;

s18 listing element number subset Q₁,Q₂,...,Q_kAll passive networks in the network are fault areas, and then fault elements are determined;

and the subscript K is the number of the element number subset obtained in the iteration process.

Further, the step S2 is specifically:

s21, constructing an initial configuration vector W based on the action information of the protection device in the SCADA system₀；

S22, judging whether the alarm signal in the power grid to be diagnosed meets the consistency of the time sequence characteristics;

if yes, go to step S23;

if not, go to step S24;

s23, using the initial configuration vector of the current structure as the final initial configuration vector W₀Proceeding to step S3;

s24, correcting the initial configuration vector through the time sequence information of the alarm signal in the power grid to be diagnosed to obtain the corrected initial configuration vector W₀The process proceeds to step S3.

Further, the initial configuration vector W in step S21₀Is represented by W₀＝(w₁₀,w₂₀,...,w_m0)^T，w_i0Is the initial object of the tissue cell, i.e. the state value, w, representing the alarm signal_i0The value of (b) is 0 or 1, i is more than or equal to 1 and less than or equal to m; when the object w is initialized_i0When the corresponding event occurs, i.e. the alarm signal corresponding to the node i is received, w_i0Setting the node state value to 1, when starting the object w_i0If the corresponding event does not occur, i.e. the alarm signal corresponding to the node i is not received, then w_i0The node state value is set to 0.

Further, in the step S22, whether the timing characteristics of the alarm signal are consistent is judged based on the timing constraint and the timing inference rule for determining the alarm signal; wherein the timing constraints of the alarm signal include a time point constraint and a time distance constraint;

the expression of the time point constraint is as follows:

wherein T (i) represents the interval of the ith event

The internal generation is carried out in the process,

and

upper and lower bounds of T (i), respectively, when

If so, the ith event is a determined event and occurs at a specific time point;

the expression of the time-distance constraint is:

in the formula, D (i, j) represents that the interval between the ith event and the jth event occurrence time must be in the time interval

In the interior of said container body,

and

respectively representing the lower and upper bounds of D (i, j).

Further, the sequential reasoning rule comprises forward sequential reasoning and reverse sequential reasoning;

wherein, the forward time sequence reasoning is as follows:

knowing the time point constraint of the ith event and the time distance constraint between the ith event and the jth event, the time point constraint of the jth event is inferred as:

the reverse time sequence reasoning is as follows:

knowing the time point constraint of the jth event and the time distance constraint of the ith event and the jth event, the time point constraint of the jth event is inferred as:

further, in step S22, the expression of the consistency of the timing characteristics determined based on the timing constraint and the timing inference rule is:

in the formula ,m_i and m_jRespectively the ith and jth events, T (m), corresponding to the alarm signal_i) Is an event m_iConstraint on the time points of occurrence, D (m)_i,m_j) Is an event m_i and m_jIs constrained.

Further, in step S3, the expression of the TTPS fault diagnosis model Π with a degree m is as follows:

Π＝(O,δ_i,E,T,i₀,syn)

wherein O is a non-null target;

δ_iis the ith histiocyte, i is more than or equal to 1 and less than or equal to m, m is the total number of the histiocytes, the histiocytes in the TTPS fault diagnosis model pi comprise real cells and virtual cells, and the forms of the histiocytes are delta_i＝(w_i0,R_i1) and δ_i＝(w_i0,R_i2)；

w_i0e.O represents an initial object multiple set in the tissue cell; r_i1Representing a transport rule of a real cell, wherein the transport rule is (i, x/lambda, j), x and lambda respectively represent objects in a cell i and a cell j, the cell i and the cell j respectively correspond to the ith event and the jth event, the object lambda in the cell j represents an empty character string, and after the rule is executed, the object x in the cell i is transmitted to the empty character string lambda in the cell j; r_i2Representing a transport rule of the virtual cells, which is in the form of (E, E/lambda, VC), wherein VC represents the virtual cells and represents main protection refusal or breaker refusal, E and lambda represent objects in the interstitial fluid environment E and the virtual cells VC, respectively, and the object lambda in the virtual cells VC represents an empty character string; after executing the rule, the object e in the interstitial fluid environment will be passed into the empty string λ in the virtual cell VC;

E＝{e₁,...,e_ndenotes the interstitial fluid environment of the tissue cells, where e_de.O, d is more than or equal to 1 and less than or equal to n, represents the objects in the interstitial fluid environment, n is the total number of the objects in the interstitial fluid environment, when no direct communication channel exists between two cells, information exchange can be indirectly carried out through the environment, and when the actual histiocyte does not meet the transport rule R_i1Then, a transport rule (i, x/λ, E) is performed, where x and λ represent objects in cell i and interstitial fluid environment E, respectively; after executing the rule, the object x in cell i will pass into the empty string λ in interstitial fluid environment E;

T＝{t_δ1,...,t_δmdenotes the set of time points for each cell, with dimensions 1 × m, m being the total number of cells, T ═ T (T) (T ═ m_δi)|δ_iE.g. delta } represents cell delta_iThe set of constraints on the point in time of,

represents cell delta_iThe occurrence time of the corresponding event should be in the time interval

Internal; d ═ D (δ)_i,δ_j)|δ_i,δ_jE δ represents a set of time-distance constraints, where

Represents cell delta_iAnd cell delta_jThe occurrence time interval of the corresponding event should be in the time interval

Internal;

i₀e { 1.. multidata, m } represents the labels of the output cells, each output cell labeled as a suspect faulty element;

the method represents the connection state set among the cells, the interconnected cells can exchange information, and all the cells (i, j) belong to syn, i is more than or equal to 1, and i is more than or equal to j when m is more than or equal to j.

Further, in step S4, the expression of the connection matrix C is set up as:

wherein ,c_ijRepresents cell delta_iAnd delta_jThere is a positive correlation between, i.e. delta_jThe corresponding event is a trigger delta_iCorresponding to the cause of the event, δ_iIs delta_jA result produced after occurrence, c_ij0 means that there is no association between two cell nodes.

Further, the judgment vector W in the step S6_endTransfer vector W₁And performing logical AND operation on the fault node vector S to construct a judgment vector W_endThe expression is as follows:

W_end＝W₁∧S

in the formula ,W₁＝(w₁₁,w₂₁,...,w_m1)^T，w_i1＝(c_i1∧w₁₁)∨(c_i2∧w₂₁)∨...∨(c_im∧w_m1)，i＝1,2,...,m；S＝(s₁,s₂,...s_n)^T，s_i∈S，s_iI is 0 or 1, 1. ltoreq. n, when s_iWhen 1 represents δ_iFor a suspect faulty element, when s_iAnd 0 represents other device elements.

The invention has the beneficial effects that:

(1) the invention introduces time sequence factors on the basis of the traditional tissue type P system, provides a time sequence tissue-like P system (TTPS), adds time point constraint to an object in cells and adds time distance constraint to a connecting channel for information exchange between the cells in a diagnosis model of the TTPS, thereby effectively utilizing time sequence information to carry out time sequence characteristic consistency constraint check between all protection devices and solving the situations of false alarm, missing alarm and the like of fault alarm signals;

(2) virtual cell and interstitial fluid environments are introduced into the TTPS, and four incidence relations which can describe action sequences and logics among various protection devices are provided on the basis, so that a new graphical modeling mode is provided, the model explanatory performance is improved, and the model complexity is reduced;

(3) the fault element is calculated through the logic operation among the initial configuration vector, the connection matrix and the fault node vector so as to deal with fault diagnosis under various fault scenes, and the fault diagnosis efficiency is improved. Therefore, by combining the time sequence characteristic of the alarm signal with the tissue type P system, the problems of false alarm, missing report and the like in the alarm signal can be well processed, the time interval of the fault occurrence time can be deduced, and the fault evolution process can be accurately described.

Drawings

Fig. 1 is a flowchart of a power grid fault diagnosis method based on a time-series organization type P system provided by the present invention.

Fig. 2 is a schematic diagram of a simple power transmission system and a TTPS fault diagnosis model provided by the present invention.

Fig. 3 is a schematic diagram of four association relationships in the TTPS fault diagnosis model provided by the present invention.

Fig. 4 is a schematic diagram of a standard IEEE14 node according to the present invention.

Fig. 5 is a schematic diagram of a TTPS fault diagnosis model of bus B13 provided by the present invention.

Detailed Description

The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.

Example 1:

the power grid fault diagnosis method based on the time sequence organization type P system provided by the embodiment of the invention comprises the following steps:

s1, determining a fault area in the power grid to be diagnosed through a junction analysis method, and further determining a suspicious fault element;

s2, constructing and correcting the initial configuration vector W based on the action information and the time sequence information of the protection device in the SCADA system₀；

S3, establishing a TTPS fault diagnosis model of each suspicious fault element according to the action sequence among the protection devices;

s4, establishing a connection matrix C and an initial configuration vector W according to the TTPS fault diagnosis model₀Constructing a translation vector

S5, generating a fault node vector S according to the node distribution of each suspicious fault element in the TTPS fault diagnosis model;

s6, according to the transfer vector W₁And a fault node vector S, and constructing a judgment vector W_endAnd according to the judgment vector W_endThe non-zero elements in the system determine the elements which send out the faults, and the fault diagnosis of the power grid is realized.

It should be noted that, in the process of executing the above steps, steps S2 and S3 are both performed based on the result obtained in step S1, and step S2 and step S3 are in parallel, and the sequence shown as S2 to S3 in this embodiment is only for forming a coherent execution logic and facilitating understanding of the solution of the present application, and in the actual implementation process, steps S2 and S3 may be executed simultaneously as needed.

Step S1 of the embodiment of the present invention specifically includes:

s11, setting the initial iteration number k to be 1;

s12, numbering each element in the power grid to be diagnosed in sequence, and constructing an element number set P_k；

S13, from set P_kIn which an arbitrary component number is taken and placed in a subset Q of component numbers_kJudging whether a closed circuit breaker exists in the power grid to be diagnosed and is connected with the currently selected element;

if yes, go to step S14;

if not, go to step S15;

s14, finding out all elements connected with the breaker in the power grid to be diagnosed, and adding the serial number of the elements to the set Q_kAnd returns to step S13;

s15, increasing the value of the search iteration number k by 1;

s16, collecting Q_k-1From the set P_k-1Get a new component number set P_k；

S17, judging the current new component number set P_kWhether it is empty;

if yes, go to step S18;

if not, returning to the step S13;

s18 listing element number subset Q₁,Q₂,...,Q_kAll passive networks in the network are fault areas, and then fault elements are determined;

and the subscript K is the number of the element number subset obtained in the iteration process.

Step S2 in the embodiment of the present invention specifically includes:

s21, constructing an initial configuration vector W based on the action information of the protection device in the SCADA system₀；

S22, judging whether the alarm signal in the power grid to be diagnosed meets the consistency of the time sequence characteristics;

if yes, go to step S23;

if not, go to step S24;

s23, using the initial configuration vector of the current structure as the final initial configuration vector W₀Proceeding to step S3;

s24, correcting the initial configuration vector through the time sequence information of the alarm signal in the power grid to be diagnosed to obtain the corrected initial configuration vector W₀The process proceeds to step S3.

In step S21, the SCADA system provides initial status information as initial access based on the TTPS fault diagnosis method, and the initial configuration vector W₀Is represented by W₀＝(w₁₀,w₂₀,...,w_m0)^T，w_i0Is the initial object of the tissue cell, i.e. the state value, w, representing the alarm signal_i0The value of (b) is 0 or 1, i is more than or equal to 1 and less than or equal to m; when the object w is initialized_i0When the corresponding event occurs, i.e. the alarm signal corresponding to the node i is received, w_i0Setting the node state value to 1, when starting the object w_i0If the corresponding event does not occur, i.e. the alarm signal corresponding to the node i is not received, then w_i0The node state value is set to 0.

In the above step S22, it is determined whether or not the timing characteristics of the alarm signal are consistent based on the timing constraint and the timing inference rule that determine the timing; wherein the timing constraints of the alarm signal include a time point constraint and a time distance constraint;

the expression of the time point constraint is as follows:

wherein T (i) represents the interval of the ith event

The internal generation is carried out in the process,

and

upper and lower bounds of T (i), respectively, when

If so, the ith event is a determined event and occurs at a specific time point;

the expression of the time-distance constraint is:

in the formula, D (i, j) represents that the interval between the ith event and the jth event occurrence time must be in the time interval

In the interior of said container body,

and

respectively representing the lower and upper bounds of D (i, j);

because the time point constraint and the time distance constraint are both defined by intervals, the time sequence characteristic constraint inference rule is defined based on the calculation rule of the intervals, and the time sequence inference rule in the embodiment comprises forward time sequence inference and reverse time sequence inference;

wherein, the forward time sequence reasoning is as follows:

knowing the time point constraint of the ith event and the time distance constraint between the ith event and the jth event, the time point constraint of the jth event is inferred as:

the reverse time sequence reasoning is as follows:

knowing the time point constraint of the jth event and the time distance constraint of the ith event and the jth event, the time point constraint of the jth event is inferred as:

based on the constraint and inference rules, the alarm signals have sequential logic association, and the sequential logic relationship can be used for judging the uncertain conditions such as the false operation of a protection device or a circuit breaker, the false alarm of the alarm signals and the like; the expression of the timing characteristic consistency determined based on the timing constraint and the timing reasoning rule is as follows:

in the formula ,m_i and m_jRespectively the ith and jth events, T (m), corresponding to the alarm signal_i) Is an event m_iConstraint on the time points of occurrence, D (m)_i,m_j) Is an event m_i and m_jIs constrained.

The meaning of the above time sequence consistency expression is: if event m_jOccurs at T (m)_j) At time, then T (m)_j) The timing characteristic consistency constraint, namely T (m), needs to be satisfied_j)∈(T(m_i)+D(m_i,m_j)). If so, event m_jTime T (m) of occurrence of_j) Completed within a specified time limit; otherwise, the corresponding alarm signal belongs to a false alarm message.

In step S3 of the present embodiment, the expression of the TTPS fault diagnosis model Π having one degree m is:

Π＝(O,δ_i,E,T,i₀,syn)

wherein O is a non-null target;

δ_iis the ith histiocyte, i is more than or equal to 1 and less than or equal to m, m is the total number of the histiocytes, the histiocytes in the TTPS fault diagnosis model pi comprise real cells and virtual cells, and the forms of the histiocytes are delta_i＝(w_i0,R_i1) and δ_i＝(w_i0,R_i2)；

w_i0e.O represents an initial object multiple set in the tissue cell; r_i1Representing a transport rule of a real cell, wherein the transport rule is (i, x/lambda, j), x and lambda respectively represent objects in a cell i and a cell j, the cell i and the cell j respectively correspond to the ith event and the jth event, the object lambda in the cell j represents an empty character string, and after the rule is executed, the object x in the cell i is transmitted to the empty character string lambda in the cell j; r_i2Representing a transport rule of the virtual cells, which is in the form of (E, E/lambda, VC), wherein VC represents the virtual cells and represents main protection refusal or breaker refusal, E and lambda represent objects in the interstitial fluid environment E and the virtual cells VC, respectively, and the object lambda in the virtual cells VC represents an empty character string; after executing the rule, the object e in the interstitial fluid environment will be passed into the empty string λ in the virtual cell VC;

E＝{e₁,...,e_ndenotes the interstitial fluid environment of the tissue cells, where e_de.O, d is more than or equal to 1 and less than or equal to n, represents the objects in the interstitial fluid environment, n is the total number of the objects in the interstitial fluid environment, when no direct communication channel exists between two cells, information exchange can be indirectly carried out through the environment, and when the actual histiocyte does not meet the transport rule R_i1Then, a transport rule (i, x/λ, E) is performed, where x and λ represent objects in cell i and interstitial fluid environment E, respectively; after executing the rule, the object x in cell i will pass into the empty string λ in interstitial fluid environment E;

T＝{t_δ1,...,t_δmdenotes the set of time points for each cell, with dimensions 1 × m, m being the total number of cells, T ═ T (T) (T ═ m_δi)|δ_iE.g. delta } represents cell delta_iThe set of constraints on the point in time of,

represents cell delta_iThe occurrence time of the corresponding event should be in the time interval

Internal; d ═ D (δ)_i,δ_j)|δ_i,δ_jE δ represents a set of time-distance constraints, where

Represents cell delta_iAnd cell delta_jThe occurrence time interval of the corresponding event should be in the time interval

Internal;

i₀e { 1.. multidata, m } represents the labels of the output cells, each output cell labeled as a suspect faulty element;

the method represents the connection state set among the cells, the interconnected cells can exchange information, and all the cells (i, j) belong to syn, i is more than or equal to 1, and i is more than or equal to j when m is more than or equal to j.

The TTPS fault diagnosis model is shown in fig. 2, and exhibits a causal relationship between a suspected fault element and a protection device by using TTPS modeling according to the action information of the protection device provided by the SCADA system. The modeling concept of the method of the present invention is illustrated below by taking as an example a simple power transmission system as shown in fig. 2(a), which includes a line L, two circuit breakers CB1 and CB2, two main protectors MLR1 and MLR2, and two backup protectors BLR1 and BLR 2. The protection of the line is configured as: if the line L has a fault, the MLR1 and the MLR2 act to trip out the circuit breakers CB1 and CB 2; if MLR1 and MLR2 fail to operate normally, then BLR1 and BLR2 will operate to trip circuit breakers CB1 and CB 2.

The fault diagnosis model of this power transmission system is established based on TTPS, as shown in fig. 2(b), in which the solid directional arcs represent causal relationships between the individual protection devices, and the time intervals on the directional arcs thereofThe connection channels with different incidence relations have different time distance constraint intervals, and the blue dotted line represents the direction of performing the inference algorithm. The model consists of 9 cells and the causal relationship logic between a suspected faulty element and its protection device is modeled by TTPS. Firstly, judging a line L in a power failure area as a suspicious element by using protection uploaded to an SCADA system and action information of a circuit breaker, wherein a cell delta corresponding to the line L₁Designated as output cell, VC (delta)₁,δ₂) Represents cell delta₁The event represented occurs, and cell delta₂The represented event does not occur.

Considering the time sequence characteristics, the tissue cells in the TTPS and the connecting channels connecting the tissue cells have the time property. And the connection channels with different association relations have different timing attributes, so four association relations of the TTPS are defined.

The four associations of TTPS are defined as shown in fig. 3:

compared with the TPS, the TTPS introduces a time point constraint and a time distance constraint to improve the fault tolerance of a diagnosis model, so that forward reasoning and time reasoning can be carried out together.

For better understanding of the TTPS model, four correlations in TTPS are described in detail. FIG. 3- (a) shows a faulty equipment node δ_jInduce a corresponding protection node delta_iAn action; FIG. 3- (b) shows the protection node δ_jTriggering the corresponding breaker cell delta_iAn action; FIG. 3- (c) shows the breaker cell delta_jFailure-induced corresponding failure-protected cell delta_iAction, VC cell represents breaker cell delta_j(ii) a rejection-triggered cell; FIG. 3- (d) shows the primary protective cell delta_jFailure to initiate the corresponding backup protective cell delta_iAction, VC cell represents the primary protective cell delta_jThe primed cells were rejected.

In the TTPS, the connection channels with different association relations have different time interval attributes, and the corresponding time distance constraint is as follows:

(1)

(failure occurrence, primary protection action) E [10,40]ms；

(2)

(protection action, breaker action) e [40,60 ]]ms；

(3)

(circuit breaker failure, failsafe action) e [210,240]m；

(4)

(Primary protection rejection, backup protection action) E [940,1030]ms. In step S4 of the present embodiment, in order to better describe the communication rule and cause-effect relationship of each cell, a connection matrix C is introduced, whose expression is:

wherein ,c_ijRepresents cell delta_iAnd delta_jThere is a positive correlation between, i.e. delta_jThe corresponding event is a trigger delta_iCorresponding to the cause of the event, δ_iIs delta_jA result produced after occurrence, c_ij0 means that there is no association between two cell nodes.

In step S4 of the present embodiment, the vector W is transferred₁Comprises the following steps:

wherein ,W₁＝(w₁₁,w₂₁,...,w_m1)^T，w_i1＝(c_i1∧w₁₁)∨(c_i2∧w₂₁)∨...∨(c_im∧w_m1) I 1, 2.. said, m; for the

Or 1, the definitions a ═ b ═ min (a, b) and a ^ b ═ max (a, b);

W₀＝(w₁₀,w₂₀,...,w_m0)^Tinitial configuration vector W constructed to represent action information of protection devices provided by SCADA system₀；

C＝(c_ij)_m×m，c _ij0 or 1, which represents the communication rule and the causal relationship of the cells corresponding to each event;

in step S5 of this embodiment, the node distribution of each faulty element in the TTPS is obtained, and a faulty node vector S is generated according to the distribution, which reflects the node distribution of each suspected faulty element in the TTPS, where S ═ S (S ═ S)₁,s₂,...s_n)^T，s_i∈S，s_iI is 0 or 1, 1. ltoreq. n, when s_iWhen 1 represents δ_iFor a suspect faulty element, when s_iWhen 0, represents other device elements;

determination vector W in step S6 of the present embodiment_endTransfer vector W₁And performing logical AND operation on the fault node vector S to construct a judgment vector W_endThe expression is as follows:

W_end＝W₁∧S

example 2:

in this embodiment, a standard IEEE14 node system is taken as a diagnosis target, and taking example 1 as an example, a specific calculation process is given to facilitate detailed understanding, and fig. 4 is a standard IEEE14 node network topology diagram.

Table 1 shows the preset failure scenario and the results of the failure diagnosis element in this embodiment

Table 1 preset fault scenario and faulty element diagnostic results of example 1

Firstly, after a fault occurs, a power failure area is judged by using a junction analysis method, and a suspected fault element bus B13, a line L1213 and a line L1314 are obtained. Establishing a TTPS fault diagnosis model of the bus B13, as shown in FIG. 5:

based on the TTPS fault diagnosis model of bus B13 of fig. 5, a 14 × 14 connection matrix C is established, which is represented as follows:

setting an initial configuration vector as W according to action information of a protection device and a breaker provided by the SCADA system₀＝(0,1,1,0,0,0,1,0,0,0,1,1,1,1)^T。

After the time sequence consistency constraint check, all the protection devices accord with the time sequence consistency constraint condition without correcting the initial configuration vector W₀＝(0,1,1,0,0,0,1,0,0,0,1,1,1,1)^T。

Connection matrix C and initial configuration vector W₀Calculating to obtain a transfer vector W₁。

Generating a fault node vector S, wherein non-0 elements in the S are a bus B13 and a virtual cell delta₁₂。

Generating a fault node vector S and a transfer vector W₁Performing logical AND operation to obtain a judgment vector W_end。

W_end＝W₁∧S＝(1,0,0,0,0,0,0,0,0,0,0,0,0,0)

The bus B13 is judged to have a fault.

And judging that the circuit L1213 and the circuit breaker CB1314 reject action in the same way.

Analyzing the fault evolution process, taking the time scale of the received first alarm as a reference point, and performing time sequence reasoning through analysis, wherein the actual fault evolution process comprises the following steps: bus B13 failed during [ -20, -10], bus main protection B13m acted at 0ms and triggered circuit breakers CB1312 and CB1306 to act at 48ms and 51ms, respectively, circuit breaker CB1314 rejected triggering far backup protection L1413s at 1989ms and tripped circuit breaker CB1413 at 2041 ms.

Claims (10)

1. A power grid fault diagnosis method based on a time sequence organization type P system is characterized by comprising the following steps:

s1, determining a fault area in the power grid to be diagnosed through a junction analysis method, and further determining a suspicious fault element;

s2, constructing and correcting the initial configuration vector W based on the action information and the time sequence information of the protection device in the SCADA system₀；

S3, establishing a TTPS fault diagnosis model of each suspicious fault element according to the action sequence among the protection devices;

s4, establishing a connection matrix C and an initial configuration vector W according to the TTPS fault diagnosis model₀Constructing a translation vector

S5, generating a fault node vector S according to the node distribution of each suspicious fault element in the TTPS fault diagnosis model;

s6, according to the transfer vector W₁And a fault node vector S, and constructing a judgment vector W_endAnd according to the judgment vector W_endThe non-zero elements in the system determine the elements which send out the faults, and the fault diagnosis of the power grid is realized.

2. The power grid fault diagnosis method based on the time-series organization type P system according to claim 1, wherein the step S1 specifically comprises:

s11, setting the initial iteration number k to be 1;

s12, numbering each element in the power grid to be diagnosed in sequence, and constructing an element number set P_k；

S13, from set P_kIn which an arbitrary component number is taken and placed in a subset Q of component numbers_kIn the method, whether the power grid to be diagnosed is in the power grid or not is judgedThe circuit breaker with the closed circuit breaker is connected with the currently selected element;

if yes, go to step S14;

if not, go to step S15;

s14, finding out all elements connected with the breaker in the power grid to be diagnosed, and adding the serial number of the elements to the set Q_kAnd returns to step S13;

s15, increasing the value of the search iteration number k by 1;

s16, collecting Q_k-1From the set P_k-1Get a new component number set P_k；

S17, judging the current new component number set P_kWhether it is empty;

if yes, go to step S18;

if not, returning to the step S13;

s18 listing element number subset Q₁,Q₂,...,Q_kAll passive networks in the network are fault areas, and then fault elements are determined;

and the subscript K is the number of the element number subset obtained in the iteration process.

3. The power grid fault diagnosis method based on the time-series organization type P system according to claim 1, wherein the step S2 specifically comprises:

s21, constructing an initial configuration vector W based on the action information of the protection device in the SCADA system₀；

S22, judging whether the alarm signal in the power grid to be diagnosed meets the consistency of the time sequence characteristics;

if yes, go to step S23;

if not, go to step S24;

s23, using the initial configuration vector of the current structure as the final initial configuration vector W₀Proceeding to step S3;

s24, correcting the initial configuration vector through the time sequence information of the alarm signal in the power grid to be diagnosed to obtain the corrected initial configuration vector W₀The process proceeds to step S3.

4. The grid fault diagnosis method based on time-series organization type P system as claimed in claim 3, wherein the initial configuration vector W in step S21₀Is represented by W₀＝(w₁₀,w₂₀,...,w_m0)^T，w_i0Is the initial object of the tissue cell, i.e. the state value, w, representing the alarm signal_i0The value of (b) is 0 or 1, i is more than or equal to 1 and less than or equal to m; when the object w is initialized_i0When the corresponding event occurs, i.e. the alarm signal corresponding to the node i is received, w_i0Setting the node state value to 1, when starting the object w_i0If the corresponding event does not occur, i.e. the alarm signal corresponding to the node i is not received, then w_i0The node state value is set to 0.

5. The grid fault diagnosis method based on time-series organization type P system as claimed in claim 3, wherein in step S22, whether the time-series characteristics are consistent is judged based on the time-series constraint and the time-series inference rule for determining the alarm signal; wherein the timing constraints of the alarm signal include a time point constraint and a time distance constraint;

the expression of the time point constraint is as follows:

wherein T (i) represents the interval of the ith event

The internal generation is carried out in the process,

and

upper and lower bounds of T (i), respectively, when

If so, the ith event is a determined event and occurs at a specific time point;

the expression of the time-distance constraint is:

in the formula, D (i, j) represents that the interval between the ith event and the jth event occurrence time must be in the time interval

In the interior of said container body,

and

respectively representing the lower and upper bounds of D (i, j).

6. The power grid fault diagnosis method based on the time sequence organization type P system as claimed in claim 5, wherein the time sequence inference rule comprises forward time sequence inference and reverse time sequence inference;

wherein, the forward time sequence reasoning is as follows:

knowing the time point constraint of the ith event and the time distance constraint between the ith event and the jth event, the time point constraint of the jth event is inferred as:

the reverse time sequence reasoning is as follows:

knowing the time point constraint of the jth event and the time distance constraint of the ith event and the jth event, the time point constraint of the jth event is inferred as:

7. the power grid fault diagnosis method based on the time-series organization type P system as claimed in claim 3, wherein in the step S22, the expression of the consistency of the time-series characteristics determined based on the time-series constraint and the time-series inference rule is as follows:

in the formula ,m_i and m_jRespectively the ith and jth events, T (m), corresponding to the alarm signal_i) Is an event m_iConstraint on the time points of occurrence, D (m)_i,m_j) Is an event m_i and m_jIs constrained.

8. The power grid fault diagnosis method based on the time-series organization type P system according to claim 1, wherein in step S3, the expression of the TTPS fault diagnosis model Π with one degree m is:

Π＝(O,δ_i,E,T,i₀,syn)

wherein O is a non-null target;

δ_iis the ith histiocyte, i is more than or equal to 1 and less than or equal to m, m is the total number of the histiocytes, the histiocytes in the TTPS fault diagnosis model pi comprise real cells and virtual cells, and the forms of the histiocytes are delta_i＝(w_i0,R_i1) and δ_i＝(w_i0,R_i2)；

w_i0e.O represents an initial object multiple set in the tissue cell; r_i1Representing the transport rule of the real cell, and the form is (i, x/lambda, j), x and lambda represent the objects in the cell i and the cell j respectively, the cell i and the cell j respectively correspond to the ith event and the jth event, and the object lambda in the cell j represents an empty characterA string, after the rule is executed, the object x in cell i will be passed into the empty string λ in cell j; r_i2Representing a transport rule of the virtual cells, which is in the form of (E, E/lambda, VC), wherein VC represents the virtual cells and represents main protection refusal or breaker refusal, E and lambda represent objects in the interstitial fluid environment E and the virtual cells VC, respectively, and the object lambda in the virtual cells VC represents an empty character string; after executing the rule, the object e in the interstitial fluid environment will be passed into the empty string λ in the virtual cell VC;

E＝{e₁,...,e_ndenotes the interstitial fluid environment of the tissue cells, where e_de.O, d is more than or equal to 1 and less than or equal to n, represents the objects in the interstitial fluid environment, n is the total number of the objects in the interstitial fluid environment, when no direct communication channel exists between two cells, information exchange can be indirectly carried out through the environment, and when the actual histiocyte does not meet the transport rule R_i1Then, a transport rule (i, x/λ, E) is performed, where x and λ represent objects in cell i and interstitial fluid environment E, respectively; after executing the rule, the object x in cell i will pass into the empty string λ in interstitial fluid environment E;

T＝{t_δ1,...,t_δmdenotes the set of time points for each cell, with dimensions 1 × m, m being the total number of cells, T ═ T (T) (T ═ m_δi)|δ_iE.g. delta } represents cell delta_iThe set of constraints on the point in time of,

represents cell delta_iThe occurrence time of the corresponding event should be in the time interval

Internal; d ═ D (δ)_i,δ_j)|δ_i,δ_jE δ represents a set of time-distance constraints, where

Represents cell delta_iAnd cell delta_jThe occurrence time interval of the corresponding event should be in the time interval

Internal;

i₀e { 1.. multidata, m } represents the labels of the output cells, each output cell labeled as a suspect faulty element;

the method represents the connection state set among the cells, the interconnected cells can exchange information, and all the cells (i, j) belong to syn, i is more than or equal to 1, and i is more than or equal to j when m is more than or equal to j.

9. The power grid fault diagnosis method based on the time-series organization type P system as claimed in claim 8, wherein in the step S4, the expression of the connection matrix C is:

wherein ,c_ijRepresents cell delta_iAnd delta_jThere is a positive correlation between, i.e. delta_jThe corresponding event is a trigger delta_iCorresponding to the cause of the event, δ_iIs delta_jA result produced after occurrence, c_ij0 means that there is no association between two cell nodes.

10. The grid fault diagnosis method based on time-series organization type P system according to claim 9, wherein the judgment vector W in the step S6_endTransfer vector W₁And performing logical AND operation on the fault node vector S to construct a judgment vector W_endThe expression is as follows:

W_end＝W₁∧S

in the formula ,W₁＝(w₁₁,w₂₁,...,w_m1)^T，w_i1＝(c_i1∧w₁₁)∨(c_i2∧w₂₁)∨...∨(c_im∧w_m1)，i＝1,2,...,m；S＝(s₁,s₂,...s_n)^T，s_i∈S，s_iI is 0 or 1, 1. ltoreq. n, when s_iWhen 1 represents δ_iFor a suspect faulty element, when s_iAnd 0 represents other device elements.

Images