CN103323707A - Transformer failure rate predicating method based on half Markoff process - Google Patents
Transformer failure rate predicating method based on half Markoff process Download PDFInfo
- Publication number
- CN103323707A CN103323707A CN2013102205321A CN201310220532A CN103323707A CN 103323707 A CN103323707 A CN 103323707A CN 2013102205321 A CN2013102205321 A CN 2013102205321A CN 201310220532 A CN201310220532 A CN 201310220532A CN 103323707 A CN103323707 A CN 103323707A
- Authority
- CN
- China
- Prior art keywords
- state
- transformer
- matrix
- platform
- transition probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Protection Of Transformers (AREA)
Abstract
The invention relates to a transformer failure rate predicating method based on the half Markoff process and belongs to the technical field of electrical equipment. According to the transformer failure rate predicating method based on the half Markoff process, the working state of a transformer is divided into an operation state and a failure rate, a thermal ageing process is set to be a progressive progress, the working state can be converted into the failure state directly from the operation state due to the influence of random factors, the converting progress of the failure rate of the transformer is irreversible, an adjacent matrix of a working state transition diagram of the transformer is obtained, the matrix of state transition probability of an embedded Markoff chain in the half Markoff process and the matrix of time condition probability distribution in the i state are obtained from historical operation data of a plurality of transformers, the matrix of transition probability in the half Markoff progress is solved with a numerical method, and the failure rate of the transformer is obtained through calculation. The transformer failure rate predicating method based on the half Markoff process has the advantages that the aging failure and the random failure are comprehensively considered, and the calculating structure is more practical.
Description
Technical field
The present invention relates to a kind of transformer fault rate Forecasting Methodology based on half Markov process, belong to the electrical equipment technical field.
Background technology
Carrying out the failure rate prediction based on accurately comprehensive transformer fault model is one of Main Means of transformer fault prevention and life-span management.The fault of transformer can be divided into degradation failure and random fault, and degradation failure is because of the performance deteriorated fault that causes gradually, such as heat ageing; Random fault refers to the fault that causes because of enchancement factor, such as human factor, thunderbolt etc.In addition, existing statistics shows in the transformer fault that causes because of short-circuit impact, and approximately 70% transformer is namely to break down behind twice short-circuit impact suffering, so short-circuit impact also can be considered random fault.
Existing transformer fault model lays particular emphasis on the research of degradation failure.The principal element that transformer is aging is heat, and regardless of the residing external environment condition of transformer, Heat Ageing runs through transformer life all the time, and is considered to affect the principal element in life-span.Therefore, heat ageing is the major consideration of existing transformer fault rate forecast model.Yet in actual motion, random fault also can not be ignored, for example short-circuit impact is the first cause of China's transformer fault, can account for 50% of total number of accident, this shows, the failure process of transformer to be described comprehensively, the factor of degradation failure and random fault two aspects need to be considered.
Markov process is powerful commonly used during equipment dependability is analyzed, existing scholar has set up the failure rate forecast model of transformer based on Markov process (can be referring to document: Ning Liaoyi, Wu Wenchuan, Zhang Baiming. become outage model () during transformer in the operation risk assessment based on the failure rate method of estimation of the inner latency fault of transformer of operating condition, Automation of Electric Systems, 2010 (15): 9-13).The document is divided into one of four states based on the volume fraction of oil dissolved gas with transformer, has set up Markov model, has obtained the variation tendency of transformer fault rate.But this model is not considered random fault, and the state of dividing based on oil dissolved gas can corresponding various faults pattern, and not pointed, practical application is difficulty relatively.In addition, necessary obeys index distribution of the residence time of Markov process claimed condition, but the physical fault process of transformer does not satisfy this requirement.
Summary of the invention
The objective of the invention is to propose a kind of transformer fault rate Forecasting Methodology based on half Markov process, set up the fault model that comprises degradation failure and random fault based on half Markov process, determine the value of Model Parameter according to statistics, to overcome the deficiency of existing Markov process model, result of calculation is conformed to the transformer fault rate curve variation tendency of actual count.
The transformer fault rate Forecasting Methodology based on half Markov process that the present invention proposes may further comprise the steps:
(1) duty with transformer is divided into running status and malfunction, running status wherein, be divided into the first state according to the degree of polymerization DP of insulating paper in the transformer: the DP value is 1100-950, the second state: the DP value is 950-450, the third state: the DP value is 450-200; Malfunction wherein is divided into according to failure mode: the 4th state: the heat ageing malfunction, and the DP value is less than 200, the 5th state: the random fault that is caused by lightning impulse, the 6th state is impacted the random fault that causes by external short circuit, above-mentioned six kinds of states form a state set S={1,2,3,4,5,6}, the duty of transformer is counted m=6;
(2) impose a condition: Heat Ageing is progressive process, and the first state, the second state and the third state are because the impact of enchancement factor can be malfunction from running state conversion directly, and the transfer process of the malfunction of transformer is irreversible,
Impose a condition according to above-mentioned, the duty conversion adjacency matrix R that obtains transformer is:
Wherein, R
Ij=1(i=1,2...6; J=1,2...6) represent to have transfer relationship, R between i state and the j state
IjThere is not transfer relationship between=0 expression i state and the j state;
(3) obtain embedded markovian state transition probability matrix P and the time conditions probability distribution matrix G (t) that rests under the i state in half Markov process from the history data of many transformers, its process may further comprise the steps:
(3-1) from the history data of many transformers, obtain respectively the moment that many transformers respectively are in running status and malfunction;
(3-2) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and definite n
3Have transfer relationship between i state and the j state among the duty of the platform transformer conversion adjacency matrix R, the transformer platform number scale that namely is transferred to the 4th state from the third state is n
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
Then embedded markovian state transition probability matrix P is in half Markov process:
(3-3) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the first state in half Markov process:
Wherein, n
12(t≤t
x) expression transfer relationship n
12Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state, t
x=1,2...T+1, T are the transformer fault rate prediction moment;
Wherein, n
15(t≤t
x) expression transfer relationship n
15Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
Wherein, n
16(t≤t
x) expression transfer relationship n
16Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the second state in half Markov process:
Wherein, n
23(t≤t
x) expression transfer relationship n
23Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
25(t≤t
x) expression transfer relationship n
25Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
26(t≤t
x) expression transfer relationship n
26Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and have transfer relationship between i state and the j state among the duty of the definite n3 platform transformer conversion adjacency matrix R is n from the transformer platform number scale that the third state is transferred to the 4th state namely
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the third state in half Markov process:
Wherein, n
34(t≤t
x) expression transfer relationship n
34Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
35(t≤t
x) expression transfer relationship n
35Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
36(t≤t
x) expression transfer relationship n
36Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Then the time conditions probability distribution matrix G (t) under the i state is in half Markov process:
(4) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in the above-mentioned Markov process, adopt numerical solution to find the solution the transition probability matrix Φ (t) of half Markov process, matrix element is expressed as Φ
Uv(t), u=1,2...m; V=1,2...m; T=1,2 ... T+1, m are transformer duty number, and t is transformer fault rate prediction current time;
Detailed process may further comprise the steps:
(4-1) during initialization, establish transformer fault rate prediction current time t=0, Q (0)=0, U (0)=E, S (0)=0, D (0)=E,
Wherein, Q (t) is half markoff kernels, and matrix element is Q
Uv(t), U (t) is the first parameter matrix, and matrix element is U
Uv(t), S (t) is the second parameter matrix, and matrix element is S
Uv(t), D (t) is the 3rd parameter matrix, and matrix element is D
Uv(t), 0 expression null matrix, E representation unit matrix;
(4-2) set t=1,
(4-3) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in above-mentioned half Markov process, obtain half markoff kernels Q (t)=PG (t), wherein the element of " " representing matrix is point-to-point multiplies each other;
(4-4) set cycle index k=1;
(4-5) according to half markoff kernels Q (t), obtain the element among the second parameter matrix S (t)
(4-6) k is judged, if k less than m, then makes k=k+1, repeating step (4-5)-(4-6); If k more than or equal to m, then carries out step (4-7);
(4-7) according to half markoff kernels Q (t) and the second parameter matrix S (t), obtain the first parameter matrix U (t)=Q (t)-Q (t-1), and the 3rd parameter matrix D (t)=D (0)-S (t);
(4-8) t is judged, if t less than T+1, then makes t=t+1, repeating step (4-3)-(4-8) is if t more than or equal to T+1, then carries out step (4-9);
When (4-9) setting t=0, Φ (0)=D (0);
(4-10) set t=1;
(4-11) according to the 3rd parameter matrix D (t), obtain transition probability matrix Φ (t)=D (t);
(4-12) set cycle index s=1;
(4-13) according to transition probability matrix and the first parameter matrix, obtain transition probability matrix Φ (t)=Φ (t)+U (s) * Φ (t-s);
(4-14) cycle index s is judged, if s less than t, s=s+1 then, repeating step (4-13)-(4-14) is if s more than or equal to t, then carries out step (4-15);
(4-15) transformer fault rate prediction current time t is judged, if t less than T+1, then makes t=t+1, repeating step (4-11)-(4-15), if t is more than or equal to T+1, then prediction finishes, and obtains the transition probability matrix Φ (t) of half Markov process;
(5) according to above-mentioned result of calculation, when obtaining transformer at failure rate prediction current time t=0 and being in the i running status, failure rate λ (t) is:
The transformer fault rate Forecasting Methodology based on half Markov process that the present invention proposes, its advantage is, only consider the deficiency of degradation failure for existing transformer fault rate model, the present invention has considered degradation failure and random fault comprehensively, computation process has adopted half Markov process, this is a kind of extended model of Markov process, the residence time that half Markov process has been cancelled running state of transformer is the restriction of exponential distribution, the residence time can be any distribution, thereby more possibilities are provided, the physical fault process that can conform to the transformer fault rate curve variation tendency of actual count is coincide better, and applicability is higher.
Embodiment
The transformer fault rate Forecasting Methodology based on half Markov process that the present invention proposes may further comprise the steps:
(1) duty with transformer is divided into running status and malfunction, running status wherein, be divided into the first state according to the degree of polymerization DP of insulating paper in the transformer: the DP value is 1100-950, the second state: the DP value is 950-450, the third state: the DP value is 450-200; Malfunction wherein is divided into according to failure mode: the 4th state: the heat ageing malfunction, and the DP value is less than 200, the 5th state: the random fault that is caused by lightning impulse, the 6th state is impacted the random fault that causes by external short circuit, above-mentioned six kinds of states form a state set S={1,2,3,4,5,6}, the duty of transformer is counted m=6;
(2) impose a condition: Heat Ageing is progressive process, and the first state, the second state and the third state are because the impact of enchancement factor can be malfunction from running state conversion directly, and the transfer process of the malfunction of transformer is irreversible,
Pass between above-mentioned six states ties up to based on analyzing under following 3 hypothesis: a) Heat Ageing is progressive process, does not consider such as the directly variation of the great-jump-forward from the first state to the third state; B) state of each normal operation: the first state, the second state and the third state, all may be because the impact of enchancement factor directly becomes malfunction from running status; C) do not consider failover process, the failure process of transformer is irreversible.
Impose a condition according to above-mentioned, the duty conversion adjacency matrix R that obtains transformer is:
Wherein, R
Ij=1(i=1,2 ... 6; J=1,2 ... 6) there is transfer relationship, R between expression i state and the j state
IjThere is not transfer relationship between=0 expression i state and the j state;
(3) obtain embedded markovian state transition probability matrix P and the time conditions probability distribution matrix G (t) that rests under the i state in half Markov process from the history data of many transformers, its process may further comprise the steps:
(3-1) from the history data of many transformers, obtain respectively the moment that many transformers respectively are in running status and malfunction;
(3-2) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
The transition probability of the second state to the six states
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and definite n
3Have transfer relationship between i state and the j state among the duty of the platform transformer conversion adjacency matrix R, the transformer platform number scale that namely is transferred to the 4th state from the third state is n
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
The transition probability of the third state to the six states
Then embedded markovian state transition probability matrix P is in half Markov process:
(3-3) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the first state in half Markov process:
Wherein, n
12(t≤t
x) expression transfer relationship n
12Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state, t
x=1,2...T+1, T are the transformer fault rate prediction moment;
Wherein, n
15(t≤t
x) expression transfer relationship n
15Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
Wherein, n
16(t≤t
x) expression transfer relationship n
16Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the second state in half Markov process:
Wherein, n
23(t≤t
x) expression transfer relationship n
23Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
25(t≤t
x) expression transfer relationship n
25Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
26(t≤t
x) expression transfer relationship n
26Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and definite n
3Have transfer relationship between i state and the j state among the duty of the platform transformer conversion adjacency matrix R, the transformer platform number scale that namely is transferred to the 4th state from the third state is n
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the third state in half Markov process:
Wherein, n
34(t≤t
x) expression transfer relationship n
34Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
35(t≤t
x) expression transfer relationship n
35Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
36(t≤t
x) expression transfer relationship n
36Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Then the time conditions probability distribution matrix G (t) under the i state is in half Markov process:
(4) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in the above-mentioned Markov process, adopt numerical solution to find the solution the transition probability matrix Φ (t) of half Markov process, matrix element is expressed as Φ
Uv(t), u=1,2...m; V=1,2...m; T=1,2 ... T+1, m are transformer duty number, and t is transformer fault rate prediction current time;
Detailed process may further comprise the steps:
(4-1) during initialization, establish transformer fault rate prediction current time t=0, Q (0)=0, U (0)=E, S (0)=0, D (0)=E,
Wherein, Q (t) is half markoff kernels, and matrix element is Q
Uv(t), U (t) is the first parameter matrix, and matrix element is U
Uv(t), S (t) is the second parameter matrix, and matrix element is S
Uv(t), D (t) is the 3rd parameter matrix, and matrix element is D
Uv(t), 0 expression null matrix, E representation unit matrix;
(4-2) set t=1,
(4-3) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in above-mentioned half Markov process, obtain half markoff kernels Q (t)=PG (t), wherein the element of " " representing matrix is point-to-point multiplies each other;
(4-4) set cycle index k=1;
(4-5) according to half markoff kernels Q (t), obtain the element among the second parameter matrix S (t)
(4-6) k is judged, if k less than m, then makes k=k+1, repeating step (4-5)-(4-6); If k more than or equal to m, then carries out step (4-7);
(4-7) according to half markoff kernels Q (t) and the second parameter matrix S (t), obtain the first parameter matrix U (t)=Q (t)-Q (t-1), and the 3rd parameter matrix D (t)=D (0)-S (t);
(4-8) t is judged, if t less than T+1, then makes t=t+1, repeating step (4-3)-(4-8) is if t more than or equal to T+1, then carries out step (4-9);
When (4-9) setting t=0, Φ (0)=D (0);
(4-10) set t=1;
(4-11) according to the 3rd parameter matrix D (t), obtain transition probability matrix Φ (t)=D (t);
(4-12) set cycle index s=1;
(4-13) according to transition probability matrix and the first parameter matrix, obtain transition probability matrix Φ (t)=Φ (t)+U (s) * Φ (t-s);
(4-14) cycle index s is judged, if s less than t, s=s+1 then, repeating step (4-13)-(4-14) is if s more than or equal to t, then carries out step (4-15);
(4-15) transformer fault rate prediction current time t is judged, if t less than T+1, then makes t=t+1, repeating step (4-11)-(4-15), if t is more than or equal to T+1, then prediction finishes, and obtains the transition probability matrix Φ (t) of half Markov process;
(5) according to above-mentioned result of calculation, when obtaining transformer at failure rate prediction current time t=0 and being in the i running status, failure rate λ (t) is:
Claims (1)
1. transformer fault rate Forecasting Methodology based on half Markov process is characterized in that this Forecasting Methodology may further comprise the steps:
(1) duty with transformer is divided into running status and malfunction, running status wherein, be divided into the first state according to the degree of polymerization DP of insulating paper in the transformer: the DP value is 1100-950, the second state: the DP value is 950-450, the third state: the DP value is 450-200; Malfunction wherein is divided into according to failure mode: the 4th state: the heat ageing malfunction, and the DP value is less than 200, the 5th state: the random fault that is caused by lightning impulse, the 6th state is impacted the random fault that causes by external short circuit, above-mentioned six kinds of states form a state set S={1,2,3,4,5,6}, the duty of transformer is counted m=6;
(2) impose a condition: Heat Ageing is progressive process, and the first state, the second state and the third state are because the impact of enchancement factor can be malfunction from running state conversion directly, and the transfer process of the malfunction of transformer is irreversible,
Impose a condition according to above-mentioned, the duty conversion adjacency matrix R that obtains transformer is:
Wherein, R
Ij=1(i=1,2...6; J=1,2...6) represent to have transfer relationship, R between i state and the j state
IjThere is not transfer relationship between=0 expression i state and the j state;
(3) obtain embedded markovian state transition probability matrix P and the time conditions probability distribution matrix G (t) that rests under the i state in half Markov process from the history data of many transformers, its process may further comprise the steps:
(3-1) from the history data of many transformers, obtain respectively the moment that many transformers respectively are in running status and malfunction;
(3-2) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
The transition probability of the first state to the second state
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and definite n
3Have transfer relationship between i state and the j state among the duty of the platform transformer conversion adjacency matrix R, the transformer platform number scale that namely is transferred to the 4th state from the third state is n
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the embedded markovian state transition probability matrix P in half Markov process:
Then embedded markovian state transition probability matrix P is in half Markov process:
(3-3) in the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of first state that is in
1, follow the tracks of this n
1Platform transformer, and definite n
1Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the first state transitions to the second state is n
12, be n from the transformer platform number scale of the first state transitions to the five states
15, be n from the transformer platform number scale of the first state transitions to the six states
16, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the first state in half Markov process:
Wherein, n
12(t≤t
x) expression transfer relationship n
12Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state, t
x=1,2...T+1, T are the transformer fault rate prediction moment;
Wherein, n
15(t≤t
x) expression transfer relationship n
15Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
Wherein, n
16(t≤t
x) expression transfer relationship n
16Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the first state;
In the moment of above-mentioned running status and malfunction, obtain the transformer number of units n of second state that is in
2, follow the tracks of this n
2Platform transformer, and definite n
2Have transfer relationship between i state and the j state among the duty conversion adjacency matrix R of platform transformer, namely the transformer platform number scale from the second state transitions to the third state is n
23, be n from the transformer platform number scale of the second state transitions to the five states
25, be n from the transformer platform number scale of the second state transitions to the six states
26, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the second state in half Markov process:
Wherein, n
23(t≤t
x) expression transfer relationship n
23Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
25(t≤t
x) expression transfer relationship n
25Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
Wherein, n
26(t≤t
x) expression transfer relationship n
26Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the second state;
In the moment of above-mentioned running status and malfunction, obtain being in the transformer number of units n of the third state
3, follow the tracks of this n
3Platform transformer, and definite n
3Have transfer relationship between i state and the j state among the duty of the platform transformer conversion adjacency matrix R, the transformer platform number scale that namely is transferred to the 4th state from the third state is n
34, the transformer platform number scale that is transferred to the 5th state from the third state is n
35, the transformer platform number scale that is transferred to the 6th state from the third state is n
36, according to above-mentioned transfer relationship, calculate the element among the time conditions probability distribution matrix G (t) under the third state in half Markov process:
Wherein, n
34(t≤t
x) expression transfer relationship n
34Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
35(t≤t
x) expression transfer relationship n
35Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Wherein, n
36(t≤t
x) expression transfer relationship n
36Be less than or equal to time t middle transfer time
xIn be in the transformer number of units of the third state;
Then the time conditions probability distribution matrix G (t) under the i state is in half Markov process:
(4) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in the above-mentioned Markov process, adopt numerical solution to find the solution the transition probability matrix Φ (t) of half Markov process, matrix element is expressed as Φ
Uv(t), u=1,2...m; V=1,2...m; T=1,2 ... T+1, m are transformer duty number, and t is transformer fault rate prediction current time;
Detailed process may further comprise the steps:
(4-1) during initialization, establish transformer fault rate prediction current time t=0, Q (0)=0, U (0)=E, S (0)=0, D (0)=E,
Wherein, Q (t) is half markoff kernels, and matrix element is Q
Uv(t), U (t) is the first parameter matrix, and matrix element is U
Uv(t), S (t) is the second parameter matrix, and matrix element is S
Uv(t), D (t) is the 3rd parameter matrix, and matrix element is D
Uv(t), 0 expression null matrix, E representation unit matrix;
(4-2) set t=1,
(4-3) according to the time conditions probability distribution matrix G (t) under the i state in embedded markovian state transition probability matrix P and half Markov process in above-mentioned half Markov process, obtain half markoff kernels Q (t)=PG (t), wherein the element of " " representing matrix is point-to-point multiplies each other;
(4-4) set cycle index k=1;
(4-5) according to half markoff kernels Q (t), obtain the element among the second parameter matrix S (t)
(4-6) k is judged, if k less than m, then makes k=k+1, repeating step (4-5)-(4-6); If k more than or equal to m, then carries out step (4-7);
(4-7) according to half markoff kernels Q (t) and the second parameter matrix S (t), obtain the first parameter matrix U (t)=Q (t)-Q (t-1), and the 3rd parameter matrix D (t)=D (0)-S (t);
(4-8) t is judged, if t less than T+1, then makes t=t+1, repeating step (4-3)-(4-8) is if t more than or equal to T+1, then carries out step (4-9);
When (4-9) setting t=0, Φ (0)=D (0);
(4-10) set t=1;
(4-11) according to the 3rd parameter matrix D (t), obtain transition probability matrix Φ (t)=D (t);
(4-12) set cycle index s=1;
(4-13) according to transition probability matrix and the first parameter matrix, obtain transition probability matrix Φ (t)=Φ (t)+U (s) * Φ (t-s);
(4-14) cycle index s is judged, if s less than t, s=s+1 then, repeating step (4-13)-(4-14) is if s more than or equal to t, then carries out step (4-15);
(4-15) transformer fault rate prediction current time t is judged, if t less than T+1, then makes t=t+1, repeating step (4-11)-(4-15), if t is more than or equal to T+1, then prediction finishes, and obtains the transition probability matrix Φ (t) of half Markov process;
(5) according to above-mentioned result of calculation, when obtaining transformer at failure rate prediction current time t=0 and being in the i running status, failure rate λ (t) is:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310220532.1A CN103323707B (en) | 2013-06-05 | 2013-06-05 | Based on the transformer fault rate Forecasting Methodology of half Markov process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310220532.1A CN103323707B (en) | 2013-06-05 | 2013-06-05 | Based on the transformer fault rate Forecasting Methodology of half Markov process |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103323707A true CN103323707A (en) | 2013-09-25 |
CN103323707B CN103323707B (en) | 2016-02-10 |
Family
ID=49192574
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310220532.1A Active CN103323707B (en) | 2013-06-05 | 2013-06-05 | Based on the transformer fault rate Forecasting Methodology of half Markov process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103323707B (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103630776A (en) * | 2013-11-19 | 2014-03-12 | 中国南方电网有限责任公司 | Evaluation method for fault probability of power equipment |
CN103714401A (en) * | 2014-01-21 | 2014-04-09 | 国家电网公司 | Method for predicting distribution and transformation fault frequency of electric network |
CN103809119A (en) * | 2013-11-26 | 2014-05-21 | 中国矿业大学 | Method for quantitative evaluation of reliability of markov model switch reluctance motor system |
CN104407267A (en) * | 2014-11-26 | 2015-03-11 | 中国船舶重工集团公司第七一二研究所 | Transformer fault detection method |
CN105093110A (en) * | 2015-09-11 | 2015-11-25 | 中国矿业大学 | Three-level Markov model switch magnetic resistance motor system reliability quantitative evaluation method |
CN105203951A (en) * | 2015-09-11 | 2015-12-30 | 中国矿业大学 | Reliability quantitative evaluation method of one-level markov model switched reluctance motor system |
CN105203952A (en) * | 2015-09-11 | 2015-12-30 | 中国矿业大学 | Method for quantitatively evaluating reliability of switched reluctance machine system through two-level Markov model |
WO2016066373A1 (en) | 2014-10-27 | 2016-05-06 | Landis+Gyr (Europe) Ag | A method, system and assembly for determining a reduction of remaining service lifetime of an electrical device during a specific time period of operation of the electrical device |
CN106779324A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | Distribution transformer state deterioration process based on Markov chain model describes method |
CN107194476A (en) * | 2017-05-26 | 2017-09-22 | 中国南方电网有限责任公司超高压输电公司天生桥局 | The pre- anti-aging maintenance policy formulating method of transformer based on semi-Markov chain |
CN107872328A (en) * | 2016-09-22 | 2018-04-03 | 华北电力大学 | Based on the polymorphic IEC61850 communication system crash rate computational methods of link |
CN108122049A (en) * | 2017-12-21 | 2018-06-05 | 郓城金河热电有限责任公司 | Photovoltaic array operation and maintenance method based on the analysis of photovoltaic module failure rate |
CN112036608A (en) * | 2020-07-31 | 2020-12-04 | 宁波伟吉电力科技有限公司 | Transformer fault detection method and detection system in electric power operation management |
CN112132195A (en) * | 2020-09-14 | 2020-12-25 | 江西山水光电科技股份有限公司 | Method for analyzing and predicting machine room fault by using Markov model |
CN112926226A (en) * | 2021-04-13 | 2021-06-08 | 西南交通大学 | Reliability evaluation method of cascade power electronic converter system |
CN114462788A (en) * | 2021-12-31 | 2022-05-10 | 浙江大学 | Multi-state element reliability evaluation method based on half Markov process |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120057513A1 (en) * | 2010-09-08 | 2012-03-08 | The Hong Kong University Of Science And Technology | Sleep mode selection employing a semi-markov decision process |
CN102436519A (en) * | 2011-08-23 | 2012-05-02 | 戴志辉 | Method for synthetically evaluating dynamic reliability of power system automatic device |
KR101163834B1 (en) * | 2009-09-29 | 2012-07-09 | 경희대학교 산학협력단 | Method for activity recognition based semi Markov conditional random fields |
CN103048573A (en) * | 2012-12-31 | 2013-04-17 | 重庆市电力公司 | Method and device for electric power system operation risk assessment |
-
2013
- 2013-06-05 CN CN201310220532.1A patent/CN103323707B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101163834B1 (en) * | 2009-09-29 | 2012-07-09 | 경희대학교 산학협력단 | Method for activity recognition based semi Markov conditional random fields |
US20120057513A1 (en) * | 2010-09-08 | 2012-03-08 | The Hong Kong University Of Science And Technology | Sleep mode selection employing a semi-markov decision process |
CN102436519A (en) * | 2011-08-23 | 2012-05-02 | 戴志辉 | Method for synthetically evaluating dynamic reliability of power system automatic device |
CN103048573A (en) * | 2012-12-31 | 2013-04-17 | 重庆市电力公司 | Method and device for electric power system operation risk assessment |
Non-Patent Citations (3)
Title |
---|
XIANG WU等: "Optimal Replacement in the Proportional Hazards Model With Semi-Markovian Covariate Process and Continuous Monitoring", 《IEEE TRANSACTIONS ON RELIABILITY》, vol. 60, no. 3, 30 September 2011 (2011-09-30), pages 580 - 589, XP011382348, DOI: doi:10.1109/TR.2011.2161049 * |
宁辽逸等: "运行风险评估中的变压器时变停运模型(二)基于延迟半马尔可夫过程的变压器时变停运模型", 《电力系统自动化》, vol. 34, no. 16, 25 August 2010 (2010-08-25), pages 24 - 28 * |
汪隆君: "电网可靠性评估方法及可靠性基础理论研究", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》, no. 06, 15 June 2011 (2011-06-15), pages 1 - 126 * |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103630776A (en) * | 2013-11-19 | 2014-03-12 | 中国南方电网有限责任公司 | Evaluation method for fault probability of power equipment |
CN103809119B (en) * | 2013-11-26 | 2016-05-18 | 中国矿业大学 | The quantitative evaluating method of Markov model switched reluctance motor system reliability |
CN103809119A (en) * | 2013-11-26 | 2014-05-21 | 中国矿业大学 | Method for quantitative evaluation of reliability of markov model switch reluctance motor system |
CN103714401A (en) * | 2014-01-21 | 2014-04-09 | 国家电网公司 | Method for predicting distribution and transformation fault frequency of electric network |
US10690729B2 (en) | 2014-10-27 | 2020-06-23 | Landis+Gyr Ag | Method, system and assembly for determining a reduction of remaining service lifetime of an electrical device during a specific time period of operation of the electrical device |
EP3447778A1 (en) | 2014-10-27 | 2019-02-27 | Landis+Gyr AG | Method, system and assembly for determining a reduction of remaining service lifetime of an electrical device based on a temperature dependent aging factor |
US10073130B2 (en) | 2014-10-27 | 2018-09-11 | Landis+Gyr Ag | Method, system and assembly for determining a reduction of remaining service lifetime of an electrical device during a specific time period of operation of the electrical device |
WO2016066373A1 (en) | 2014-10-27 | 2016-05-06 | Landis+Gyr (Europe) Ag | A method, system and assembly for determining a reduction of remaining service lifetime of an electrical device during a specific time period of operation of the electrical device |
CN104407267B (en) * | 2014-11-26 | 2017-10-20 | 中国船舶重工集团公司第七一二研究所 | A kind of transformer fault detection method |
CN104407267A (en) * | 2014-11-26 | 2015-03-11 | 中国船舶重工集团公司第七一二研究所 | Transformer fault detection method |
WO2017041391A1 (en) * | 2015-09-11 | 2017-03-16 | 中国矿业大学 | Method of performing quantitative assessment on reliability of switching reluctance motor system using first-order markov model |
WO2017041392A1 (en) * | 2015-09-11 | 2017-03-16 | 中国矿业大学 | Method of performing quantitative assessment on reliability of switching reluctance motor system using second-order markov model |
WO2017041393A1 (en) * | 2015-09-11 | 2017-03-16 | 中国矿业大学 | Method of performing quantitative assessment on reliability of switching reluctance motor system using third-order markov model |
CN105203952A (en) * | 2015-09-11 | 2015-12-30 | 中国矿业大学 | Method for quantitatively evaluating reliability of switched reluctance machine system through two-level Markov model |
CN105203951A (en) * | 2015-09-11 | 2015-12-30 | 中国矿业大学 | Reliability quantitative evaluation method of one-level markov model switched reluctance motor system |
CN105093110A (en) * | 2015-09-11 | 2015-11-25 | 中国矿业大学 | Three-level Markov model switch magnetic resistance motor system reliability quantitative evaluation method |
CN107872328A (en) * | 2016-09-22 | 2018-04-03 | 华北电力大学 | Based on the polymorphic IEC61850 communication system crash rate computational methods of link |
CN107872328B (en) * | 2016-09-22 | 2021-04-02 | 华北电力大学 | Method for calculating IEC61850 communication system failure rate based on multiple states of communication link |
CN106779324A (en) * | 2016-11-29 | 2017-05-31 | 武汉大学 | Distribution transformer state deterioration process based on Markov chain model describes method |
CN107194476A (en) * | 2017-05-26 | 2017-09-22 | 中国南方电网有限责任公司超高压输电公司天生桥局 | The pre- anti-aging maintenance policy formulating method of transformer based on semi-Markov chain |
CN107194476B (en) * | 2017-05-26 | 2019-12-10 | 中国南方电网有限责任公司超高压输电公司天生桥局 | aging prevention maintenance strategy making method of transformer based on half Markov chain |
CN108122049A (en) * | 2017-12-21 | 2018-06-05 | 郓城金河热电有限责任公司 | Photovoltaic array operation and maintenance method based on the analysis of photovoltaic module failure rate |
CN112036608A (en) * | 2020-07-31 | 2020-12-04 | 宁波伟吉电力科技有限公司 | Transformer fault detection method and detection system in electric power operation management |
CN112132195A (en) * | 2020-09-14 | 2020-12-25 | 江西山水光电科技股份有限公司 | Method for analyzing and predicting machine room fault by using Markov model |
CN112132195B (en) * | 2020-09-14 | 2024-03-29 | 江西山水光电科技股份有限公司 | Method for analyzing and predicting machine room faults by using Markov model |
CN112926226A (en) * | 2021-04-13 | 2021-06-08 | 西南交通大学 | Reliability evaluation method of cascade power electronic converter system |
CN114462788A (en) * | 2021-12-31 | 2022-05-10 | 浙江大学 | Multi-state element reliability evaluation method based on half Markov process |
CN114462788B (en) * | 2021-12-31 | 2023-10-31 | 浙江大学 | Multi-state element reliability assessment method based on semi-Markov process |
Also Published As
Publication number | Publication date |
---|---|
CN103323707B (en) | 2016-02-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103323707B (en) | Based on the transformer fault rate Forecasting Methodology of half Markov process | |
CN103646358B (en) | Meter and the electrical network scheduled overhaul cycle determination method of power equipment time-varying fault rate | |
CN104537487B (en) | A kind of appraisal procedure of power transmission and transforming equipment operation state risk | |
CN104217104A (en) | Power transformer service life analysis method and system based on risk evaluation | |
CN106908674A (en) | A kind of Transformer condition evaluation based on the prediction of multimode amount | |
CN110287543B (en) | Method for predicting service life of relay protection device | |
Abbasghorbani et al. | Reliability‐centred maintenance for circuit breakers in transmission networks | |
WO2016004361A1 (en) | System and method for assessing smart power grid networks | |
CN110502777A (en) | IGBT module condition detecting system and method based on neural network prediction | |
CN106447201A (en) | Optimal maintenance decision method of power transmission and transformation equipment based on Markov decision process | |
CN115936448A (en) | Urban distribution network power evaluation system and method based on big data | |
CN105069701A (en) | Monte Carlo method based risk evaluation method for power transmission system | |
Liu et al. | Reinforcement learning for cyber-physical security assessment of power systems | |
CN103985059B (en) | Power grid transformer operational risk assessment method based on fuzzy fault rate | |
CN104680239A (en) | Distribution network maintenance scheme optimization method and device based on element failure model | |
Huang et al. | Vulnerability analysis of electrical cyber physical systems using a simulation platform | |
CN105067877A (en) | Method of calculating power consumption mean value of multiple transformer parallel power supply system | |
CN104008509A (en) | Method and device for analysis of power supply reliability calculation | |
CN104036131A (en) | Transformer aging fault rate estimation method | |
CN105868544A (en) | A method of analyzing insulating material performance by using three-parameter Weibull distribution to process flashover voltages | |
CN111443305A (en) | Method and device for detecting line fault by using electric energy deviation value | |
CN105307202A (en) | Method for evaluating availability of distribution automation wireless communication | |
CN111985708B (en) | Bayesian algorithm-based high-voltage circuit breaker refusal operation probability prediction method and device | |
Guo et al. | Reliable life calculation of circuit breaker based on FTA and reliability function | |
Wang | Research on Fault Predition and Self-healing Control Technology of Distribution Network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |