Failure analysis method for aluminum cable steel reinforced of electric power system based on image processing

Abstract

The invention belongs to the technical field of electric power fault identification, and particularly relates to an electric power system steel-cored aluminum strand failure analysis method based on image processing, which comprises the following steps: acquiring wire surface data information; acquiring internal data information of the wire; the wire surface data information and the wire internal data information are sent to an analysis terminal; comparing the data with stored data in each characteristic standard database to obtain each characteristic failure early warning threshold value; according to the method provided by the invention, when the lead fails, the failure type of the lead can be quickly judged, the accident reason can be quickly found out, the cost is low, the detection efficiency is high, and a feasible analysis scheme is provided for different environments.

Description

Technical Field

[0001] The present application belongs to the technical field of power fault identification, and specifically relates to a method for analyzing failure of steel-core aluminum stranded wire in a power system based on image processing. Background Art

[0002] Overhead conductors are the main body of high-voltage transmission lines. At present, overhead conductors of high-voltage transmission and distribution networks at home and abroad are mainly steel-core aluminum stranded conductors. Overhead conductors have been in service for 25-40 years and are in the middle of their lifespan. Since overhead conductors are exposed to the open air for a long time, they are subject to harsh natural conditions such as wind, rain, ice and snow, and are prone to various accidents. The maintenance and renewal tasks of in-service overhead conductors are arduous. How to scientifically evaluate the operating conditions of in-service overhead conductors and reasonably predict their remaining lifespan is very important, and has become a research hotspot in power transmission engineering and related fields.

[0003] The failure of conductor fracture is the biggest enemy of power grid security, causing great economic losses to the power transmission of the power grid, and causing great impact on the primary and secondary equipment of the power system. In addition, replacing the conductor requires a lot of time, manpower and material resources. At present, there is no perfect supervision and inspection method for conductors, which is limited to appearance inspection, component analysis and other tests. With the increase of voltage level, the requirements for conductor current carrying and force are greatly improved, and the traditional detection mode is gradually unable to meet the needs. Through comprehensive analysis and testing of various types of conductors, the surface characteristics, organizational characteristics, mechanical properties, material aging, corrosion, fatigue and other characteristics of the conductors are mastered, and an evaluation model for the life of the conductor is formed, so that the supervision and inspection work can be carried out in a targeted manner.

[0004] Steel core aluminum stranded wire conductor is a composite structure with inner and outer aluminum wires and galvanized steel core wires twisted together. After atmospheric corrosion and fretting wear, the mechanical and electrical properties of the conductor are reduced. After years of accumulation, it may lead to line breakage, line breakage, and even power outages, causing immeasurable losses. The damage of this type of conductor mainly comes from: First, the quality defects of the conductor lead to early failure, mainly oil stains on the surface of the conductor, over-twisting of the conductor, single-wire breakage, surface scratches of single wires or conductors, single-wire riding (jumping wires), single-wire arching, lantern-shaped conductors, serpentine bends and loose strands; second, corrosion damage to the conductor; third, fretting damage to the conductor, mainly dancing damage to the conductor, sub-blocking structure characteristics, mechanical properties, material aging, corrosion, fatigue characteristic changes of the conductor, and breeze structure characteristics, mechanical properties, material aging, corrosion, fatigue characteristic changes of the conductor; fourth, creep damage to the conductor.

[0005] Conductors break and fail due to problems in design, manufacturing, installation, operation, maintenance, and management, as well as long-term operation under harsh conditions such as scorching sun, ice and snow, strong winds, oil pollution, and seawater. Therefore, by formulating technical plans and conducting various tests, we study the changes in conductor performance, summarize the failures caused by quality defects in conductors, conduct failure analysis on conductors with different failure forms, determine their failure causes, and prevent the recurrence of failure, which has great economic benefits for ensuring the safety of the power grid. Summary of the invention

[0006] In order to solve the problems existing in the prior art, the present application proposes a failure analysis method of steel-core aluminum stranded wire in a power system based on image processing.

[0007] To achieve the above objectives, the technical solutions adopted by this application are as follows:

[0008] A method for analyzing failure of steel-core aluminum stranded wire in a power system based on image processing comprises the following steps:

[0009] S1: Use the camera equipment mounted on the cruise drone to cyclically acquire 360° images of the conductor surface to obtain surface data information;

[0010] S2: Use the non-destructive detector carried on the cruise drone to obtain the internal data information of the conductor;

[0011] S3: sending the conductor surface data information and the conductor internal data information obtained above to the power system steel core aluminum stranded wire failure analysis terminal via wireless signals;

[0012] S4: Compare the data with the data in the standard database of each feature of the intact conductor stored to obtain the failure warning threshold of each feature;

[0013] S5: The acquired failure warning thresholds of each feature are sent to the power system steel core aluminum stranded wire failure monitoring center for visual display.

[0014] Furthermore, the conductor surface data information in step S1 includes surface integrity characteristic data of the conductor.

[0015] Furthermore, the internal data information of the wire in step S2 includes tissue characteristic data, mechanical property data, material aging data, corrosion data, and fatigue characteristic data.

[0016] Furthermore, the method of cyclically acquiring the surface data information of the conductor using the camera device mounted on the cruise drone includes the following steps:

[0017] S101: Data input, including the high-definition camera matrix H, the surface integrity vector p of the wire, the maximum number of iterations m, the threshold parameter γ, the inversion parameter β, and the optimal convenience matrix Y;

[0018] S102: Data initialization, let the initial deviation r 0 =p, the surface integrity base set of the primary selected conductor Basic set of surface integrity of final conductors The number of iterations s = 1; the initial optimal convenience matrix γ is an N×N identity matrix; is an empty set, N is the pixel points on the wire surface;

[0019] S103: Update the optimal convenience matrix Y, the surface integrity basis set ∧ sk The corresponding position value is 0, and the other position values ​​are 1;

[0020] S104: updating the high-definition camera matrix H, H=Γ×Z, where Γ is the surface integrity matrix of the wire, which is used to perceive the brightness of the original signal represented by the image grayscale, and Z is the image grayscale matrix, which is used to represent the original signal by the image grayscale;

[0021] S105: Pre-selected set of surface integrity of primary selected conductors, expressed as:

[0022] u=abs[H Y r s-1 ]+ <r s-1 , a j >

[0023] Where u is the preselected set of conductor surface integrity, abs[] is the preselected function, and H Y To update the high-definition camera matrix under the optimal convenient matrix constraint, r s-1 is the deviation of s-1 iterations, a j is the j-th column of the matrix H, <r s-1 , a j >In order to seek s-1 , a j The inner product of

[0024] S106: Update the pre-selected set of surface integrity of the primary selected conductors, the expression is:

[0025] ∧ s =∧ s-1 ∪J 0

[0026] H s =H s-1 ∪a j

[0027] In the formula, ∧ s is the surface integrity basis set, ∧ s-1 is the surface integrity basis set before updating, J 0is the serial number of matrix H, H s is the surface integrity of the preliminary selected wire obtained in the sth iteration, H s-1 is the surface integrity of the preliminary selected wire obtained in the s1th iteration;

[0028] S107: Find the minimum optimal solution, the expression is:

[0029]

[0030] In the formula, is the minimum optimization solution value, arg min|| || is the minimum optimization solution function, D s is the optimal solution of s iterations, is the HD camera matrix under the constraint of updating the optimal convenient matrix in s iterations, ρ is the sparsity of the camera points;

[0031] S108: Update the surface integrity pre-selection set of the final selected conductor, from The w items with the largest absolute value are selected from the final selected w conductor surface integrity pre-selection set, corresponding to the preliminary selected conductor surface integrity pre-selection set H s The W column in is denoted as H sw ;

[0032] s109: Update deviation, the expression is:

[0033]

[0034] In the formula, r s is the deviation under s iterations;

[0035] S110: If the number of iterations does not reach the maximum number of iterations, set s=s+1 and return to step S103 to continue iterating. If the number of iterations reaches the maximum number of iterations or the deviation, set s=0 and stop iterating and proceed to step S111;

[0036] S111: Output an approximate value of the reconstructed image information of the surface integrity of the wire.

[0037] Furthermore, in step S4, the data in the stored standard database of each characteristic of the intact wire is compared to obtain the failure warning threshold of each characteristic, which includes the following steps:

[0038] S401, state model selection: according to the range of the failure warning threshold of each feature obtained above, it is divided into an obvious state model and a weak state model; the obvious state model is the data obtained by the camera equipment on the cruise drone that the surface damage of the wire exceeds 10%, and the data obtained by the non-destructive detector that the tissue characteristics inside the wire change, the mechanical properties are less than 50%, the corrosion rate exceeds 30%, and the fatigue characteristics exceed 20%; the weak state model is the data information opposite to the obvious state model.

[0039] S402, measurement model selection: perform Kalman filter calculation for the obvious state model, and perform model probability calculation and model mixing processing for the weak state model.

[0040] Furthermore, in step S402, Kalman filter calculation is performed on the obvious state model, and its expression is:

[0041] X(k)=Φ cp X(k-1)+w cp (k)

[0042] In the formula, X(k) is the state parameter at the current moment, X(k-1) is the state parameter at the previous moment, Φ cp is the state transfer matrix of the normal operation state model, w cp (k) is the process vector of the normal operation state model, k is the kth elapsed time period;

[0043] Under the obvious state model, the state transfer matrix Φ of the normal operating state model cp The value of is:

[0044]

[0045] Furthermore, in step S402, the model probability calculation is performed for the weak state model, and the expression is:

[0046]

[0047] In the formula, is the parameter change acceleration measured under the weak state model, w 0 is the process vector of the initial state model, is the process vector coefficient of the initial state model, x is the initial acceleration of the weak state model, It is the initial acceleration when the weak state model changes.

[0048] Furthermore, in step S402, the weak state model is subjected to model mixing processing, including

[0049] (1) Using the model probabilities λ of the previous time period i(k-1), state failure analysis value and the Markov chain π between models ij ;

[0050]

[0051] In the formula, i is the model of the previous time period, l is the total amount of models in all time periods, and λ j is the probability of the jth column, k is the elapsed time period;

[0052] (2) Based on the model probability and model probability of the previous time period, calculate the contribution λ of model i to model j in the previous time period i|j , the expression is:

[0053]

[0054] In the formula, λ i|j is the contribution of model i to model j, is the inversion contribution of the j-th column probability;

[0055] (3) Calculation of state failure analysis value The expression is:

[0056]

[0057] In the formula, is the state failure analysis value of the model during the previous elapsed time period;

[0058] State Failure Analysis Value The covariance matrix Q(k|k) of is calculated as follows:

[0059]

[0060] In the formula, Q i (k|k) is the covariance matrix, is the state deviation analysis value of the model over time.

[0061] (4) The judgment criteria for obtaining the failure warning thresholds of the above-mentioned characteristics are: the value calculated by the covariance matrix Q(k|k) of the final state failure analysis value obtained is greater than 10% as the failure warning threshold of each characteristic.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] This application uses image processing methods to quickly detect the organizational characteristic data, mechanical property data, material aging data, corrosion data, and fatigue characteristic data of the wire. The procedure is simple, the cost is low, and the detection efficiency is high. It provides a practical analysis plan for different environments. The above test results are combined to analyze the cause of the wire failure, draw conclusions and issue failure analysis results. This application can also be combined with later inspection and analysis tests, which are divided into macroscopic inspection, component analysis, electrochemical corrosion analysis, electron microscopy and energy spectrum analysis, resistivity test and tensile test; this application method is used as a guide for supervision and inspection work. When the wire fails, it can quickly identify its failure type, find out the cause of the accident, and improve efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A flowchart of the method for failure analysis of steel-core aluminum stranded wire in power system based on image processing provided in this application; DETAILED DESCRIPTION

[0065] The technical solution of the present application will be clearly and completely described below in conjunction with the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments; based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0066] like Figure 1 As shown, the present application proposes a failure analysis method of steel-core aluminum stranded wire in a power system based on image processing, comprising the following steps:

[0067] S1: Use the camera equipment carried on the cruise drone to cyclically obtain 360° images of the conductor surface to obtain surface data information, including the surface integrity characteristic data of the conductor;

[0068] S2: Use the non-destructive detector carried on the cruise drone to obtain the internal data information of the conductor, including organizational characteristics data, mechanical properties data, material aging data, corrosion data, and fatigue characteristic data;

[0069] S3: sending the conductor surface data information and the conductor internal data information obtained above to the power system steel core aluminum stranded wire failure analysis terminal via wireless signals;

[0070] S4: Compare the data with the data in the standard database of each feature of the intact conductor stored to obtain the failure warning threshold of each feature;

[0071] S5: The acquired failure warning thresholds of each feature are sent to the power system steel core aluminum stranded wire failure monitoring center for visual display.

[0072] In this application, the surface data information of the conductor is cyclically acquired by using a camera device mounted on a cruise drone, including the following steps:

[0073] S101: Data input, including the high-definition camera matrix H, the surface integrity vector p of the wire, the maximum number of iterations m, the threshold parameter γ, the inversion parameter β, and the optimal convenience matrix Y;

[0074] S102: Data initialization, let the initial deviation r 0 =p, the surface integrity base set of the primary selected conductor Basic set of surface integrity of final conductors The number of iterations s = 1; the initial optimal convenience matrix Y is an N×N identity matrix; is an empty set, N is the pixel points on the wire surface;

[0075] S103: Update the optimal convenience matrix Y, the position value corresponding to the surface integrity basis set ∧sk is 0, and the other position values ​​are 1;

[0076] S104: updating the high-definition camera matrix H, H=Γ×Z, where Γ is the surface integrity matrix of the wire, which is used to perceive the brightness of the original signal represented by the image grayscale, and Z is the image grayscale matrix, which is used to represent the original signal by the image grayscale;

[0077] S105: Pre-selected set of surface integrity of primary selected conductors, expressed as:

[0078] u=abs[H Y r s-1 ]+ <r s-1 , a j >

[0079] Where u is the preselected set of conductor surface integrity, abbs[] is the preselected function, and H Y To update the high-definition camera matrix under the optimal convenient matrix constraint, r s-1 is the deviation of s-1 iterations, a j is the j-th column of the matrix H, <r s-1 , a j >In order to seek s-1 , a j The inner product of

[0080] S106: Update the pre-selected set of surface integrity of the primary selected conductors, the expression is:

[0081] ∧ s =∧ s-1 ∪J 0

[0082] H s =Hs-1 ∪a j

[0083] In the formula, ∧ s is the surface integrity basis set, ∧ s-1 is the surface integrity basis set before updating, J 0 is the serial number of matrix H, H s is the surface integrity of the preliminary selected wire obtained in the sth iteration, H s-1 is the surface integrity of the preliminary selected conductor obtained in the s-1th iteration;

[0084] S107: Find the minimum optimal solution, the expression is:

[0085]

[0086] In the formula, is the minimum optimization solution value, arg min|| || is the minimum optimization solution function, D s is the optimal solution of s iterations, is the HD camera matrix under the constraint of updating the optimal convenient matrix in s iterations, ρ is the sparsity of the camera points;

[0087] S108: Update the surface integrity pre-selection set of the final selected conductor, from The w items with the largest absolute value are selected from the final selected w conductor surface integrity pre-selection set, corresponding to the preliminary selected conductor surface integrity pre-selection set H s The W column in is denoted as H sw ;

[0088] S109: Update deviation, the expression is:

[0089]

[0090] In the formula, r s is the deviation under s iterations;

[0091] S110: If the number of iterations does not reach the maximum number of iterations, set s=s+1 and return to step S103 to continue iterating. If the number of iterations reaches the maximum number of iterations or the deviation, set s=0 and stop iterating and proceed to step S111;

[0092] s111: Outputs an approximate value of the reconstructed image information of the surface integrity of the wire.

[0093] In step S4 of the present application, the data in the stored standard database of each characteristic of the intact wire is compared to obtain the failure warning threshold of each characteristic, including the following steps:

[0094] S401, state model selection: according to the range of the failure warning threshold of each feature obtained above, it is divided into an obvious state model and a weak state model; the obvious state model is the data obtained by the camera equipment on the cruise drone that the surface damage of the wire exceeds 10%, and the data obtained by the non-destructive detector that the tissue characteristics inside the wire change, the mechanical properties are less than 50%, the corrosion rate exceeds 30%, and the fatigue characteristics exceed 20%; the weak state model is the data information opposite to the obvious state model.

[0095] S402, measurement model selection: Kalman filter calculation is performed for the obvious state model, and model probability calculation and model mixing processing are performed for the weak state model;

[0096] For the obvious state model, the Kalman filter calculation is performed and the expression is:

[0097] X(k)=Φ cp X(k-1)+w cp (k)

[0098] In the formula, X(k) is the state parameter at the current moment, X(k-1) is the state parameter at the previous moment, Φ cp is the state transfer matrix of the normal operation state model, w cp (k) is the process vector of the normal operation state model, k is the kth elapsed time period;

[0099] Under the obvious state model, the state transfer matrix Φ of the normal operating state model cp The value of is:

[0100]

[0101] For the weak state model, the model probability calculation is performed, and its expression is:

[0102]

[0103] In the formula, is the parameter change acceleration measured under the weak state model, w 0 is the process vector of the initial state model, is the process vector coefficient of the initial state model, x is the initial acceleration of the weak state model, It is the initial acceleration when the weak state model changes.

[0104] Model mixing processing is performed for the weak state model, including:

[0105] (1) Using the model probabilities λ of the previous time period i (k-1), state failure analysis value and the Markov chain π between modelsij ;

[0106]

[0107] In the formula, i is the model of the previous time period, l is the total amount of models in all time periods, and λ j is the probability of the jth column, k is the elapsed time period;

[0108] (2) Based on the model probability and model probability of the previous time period, calculate the contribution λ of model i to model j in the previous time period i|j , the expression is:

[0109]

[0110] In the formula, λ i|j is the contribution of model i to model j, is the inversion contribution of the j-th column probability;

[0111] (3) Calculation of state failure analysis value The expression is:

[0112]

[0113] In the formula, is the state failure analysis value of the model during the previous elapsed time period;

[0114] State Failure Analysis Value The covariance matrix Q(k|k) of is calculated as follows:

[0115]

[0116] In the formula, Q i (k|k) is the covariance matrix, is the state deviation analysis value of the model over time.

[0117] (4) The judgment criteria for obtaining the failure warning thresholds of the above-mentioned characteristics are: the value calculated by the covariance matrix Q(k|k) of the final state failure analysis value obtained is greater than 10% as the failure warning threshold of each characteristic.

[0118] The above description is only a preferred specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by any technician familiar with the technical field within the technical scope disclosed in the present application and within the spirit and principle of the present application shall be covered by the protection scope of the present application.

Claims

1. A method for failure analysis of steel-core aluminum stranded wire in power system based on image processing, comprising the following steps: S1: Use the camera equipment mounted on the cruise drone to cyclically acquire 360° images of the conductor surface to obtain surface data information; S2: Use the non-destructive detector carried on the cruise drone to obtain the internal data information of the conductor; S3: sending the conductor surface data information and the conductor internal data information obtained above to the power system steel core aluminum stranded wire failure analysis terminal via wireless signals; S4: Compare the data with the data in the standard database of each feature of the intact conductor stored to obtain the failure warning threshold of each feature; S5: The acquired failure warning thresholds of each feature are sent to the power system steel core aluminum stranded wire failure monitoring center for visual display.

2. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 1 is characterized in that: The conductor surface data information in step S1 includes surface integrity characteristic data of the conductor.

3. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 1 is characterized in that: The internal data information of the wire in step S2 includes tissue characteristic data, mechanical property data, material aging data, corrosion data, and fatigue characteristic data.

4. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 1, characterized in that: The method of cyclically acquiring the surface data information of the conductor using the camera device mounted on the cruise drone includes the following steps: S101: Data input, including the high-definition camera matrix H, the surface integrity vector p of the wire, the maximum number of iterations m, the threshold parameter γ, the inversion parameter β, and the optimal convenience matrix Y; S102: Data initialization, set the initial deviation r0 = p, and preliminarily select the basic set of surface integrity of the conductor Basic set of surface integrity of final conductors The number of iterations s = 1; the initial optimal convenience matrix Y is an N×N identity matrix; is an empty set, N is the pixel points on the wire surface; S103: Update the optimal convenience matrix Y, the surface integrity basis set ∧ sk The corresponding position value is 0, and the other position values ​​are 1; S104: updating the high-definition camera matrix H, H=Γ×Z, where Γ is the surface integrity matrix of the wire, which is used to perceive the brightness of the original signal represented by the image grayscale; Z is the image grayscale matrix, which is used to represent the original signal by the image grayscale; S105: Pre-selected set of surface integrity of primary selected conductors, expressed as: u=abs[H Y r s-1 ]+<r s-1 ,a j > Where u is the preselected set of conductor surface integrity, abs[] is the preselected function, and H Y To update the high-definition camera matrix under the optimal convenient matrix constraint, r s-1 is the deviation of s-1 iterations, a j is the j-th column of the matrix H, <r s-1 , a j >In order to seek s-1 , a j The inner product of S106: Update the pre-selected set of surface integrity of the primary selected conductors, the expression is: ∧ s ＝∧ s-1 ∪J0 H s =H s-1 ∪a j In the formula, ∧ s is the surface integrity basis set, ∧ s-1 is the surface integrity basis set before updating, J0 is the serial number of matrix H, H s is the surface integrity of the preliminary selected wire obtained in the sth iteration, H s-1 is the surface integrity of the preliminary selected conductor obtained in the s-1th iteration; S107: Find the minimum optimal solution, the expression is: In the formula, is the minimum optimization solution value, arg min|||| is the minimum optimization solution function, D s is the optimal solution of s iterations, is the HD camera matrix under the constraint of updating the optimal convenient matrix in s iterations, ρ is the sparsity of the camera points; S108: Update the surface integrity pre-selection set of the final selected conductor, from The W items with the largest absolute value are selected from the final W wire surface integrity pre-selection set, corresponding to the preliminary wire surface integrity pre-selection set H s The W column in is denoted as H sW ; S109: Update deviation, the expression is: In the formula, r s is the deviation under s iterations; S110: If the number of iterations does not reach the maximum number of iterations, set s=s+1 and return to step S103 to continue iterating. If the number of iterations reaches the maximum number of iterations or the deviation, set s=0 and stop iterating and proceed to step S111; S111: Output an approximate value of the reconstructed image information of the surface integrity of the wire.

5. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 1, characterized in that: In step S4, the data in the stored standard database of each characteristic of the intact conductor are compared to obtain the failure warning threshold of each characteristic, including the following steps: S401, state model selection: according to the range of the failure warning threshold of each feature obtained above, the state model is divided into an obvious state model and a weak state model; S402, measurement model selection: perform Kalman filter calculation for the obvious state model, and perform model probability calculation and model mixing processing for the weak state model.

6. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 5 is characterized in that: In step S401, the obvious state model is data obtained by the camera equipment on the cruise drone showing that the surface damage of the wire exceeds 10%, and the non-destructive detector shows that the tissue characteristics inside the wire change, the mechanical properties are less than 50%, the corrosion rate exceeds 30%, and the fatigue characteristics exceed 20%; the weak state model is data information opposite to the obvious state model.

7. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 5, characterized in that: In step S402, Kalman filter calculation is performed on the obvious state model, and its expression is: X(k)=Φ cp X(k-1)+w cp (k) In the formula, X(k) is the state parameter at the current moment, X(k-1) is the state parameter at the previous moment, Φ cp is the state transfer matrix of the normal operation state model, w cp (k) is the process vector of the normal operation state model, k is the kth elapsed time period; Under the obvious state model, the state transfer matrix Φ of the normal operating state model cp The value of is:

8. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 5 is characterized by: In step S402, the model probability calculation is performed for the weak state model, and the expression is: In the formula, is the parameter change acceleration measured under the weak state model, w0 is the process vector of the initial state model, ζ is the process vector coefficient of the initial state model, x is the initial acceleration of the weak state model, It is the initial acceleration when the weak state model changes.

9. The method for analyzing failure of steel-core aluminum stranded wire in power system based on image processing according to claim 5, characterized in that: In step S402, the weak state model is subjected to model mixing processing, including: (1) Using the model probabilities λ of the previous time period i (k-1), state failure analysis value and the Markov chain π between models ij ; In the formula, i is the model of the previous time period, l is the total amount of models in all time periods, and λ j is the probability of the jth column, k is the elapsed time period; (2) Based on the model probability and model probability of the previous time period, calculate the contribution λ of model i to model j in the previous time period i|j , the expression is: In the formula, λ i|j is the contribution of model i to model j, is the inversion contribution of the j-th column probability; (3) Calculation of state failure analysis value The expression is: In the formula, is the state failure analysis value of the model during the previous elapsed time period; State Failure Analysis Value The covariance matrix Q(k|k) of is calculated as follows: In the formula, Q i (k|k) is the covariance matrix, is the state deviation analysis value of the model over time. (4) The judgment criteria for obtaining the failure warning thresholds of the above-mentioned characteristics are: the value calculated by the covariance matrix Q(k|k) of the final state failure analysis value obtained is greater than 10% as the failure warning threshold of each characteristic.

Images