A comprehensive availability analysis method for power communication optical cables considering service characteristics

Abstract

The present invention discloses a comprehensive availability analysis method for power communication optical cables taking into account business characteristics, including: extracting the characteristics of the business carried by the optical cable and its values, and calculating the business volume; extracting the characteristics of the optical cable and its values, and calculating the availability of the optical cable; and calculating the remaining proportion of the fiber core resources. Using the business volume, optical cable availability and the remaining proportion of the fiber core resources, extreme data objects are cleaned and an analysis data set is constructed; a difference matrix is established; the number of clusters is estimated using a clustering trend visualization evaluation algorithm; a fuzzy clustering algorithm is cyclically executed to obtain an optimized clustering result; and the availability of power communication optical cable resources in each cluster is analyzed through data visualization technology. The present invention combines business and optical cable characteristics to comprehensively solve the problem of rational utilization of power communication optical cable resources, helps to accurately identify the business carrying status of the optical cable and the availability of the optical cable resources, rationally plan and deploy power communication services, and reduce the risk of optical cable load.

Description

Technical Field

[0001] The present invention relates to the technical field of power communication, and in particular to a comprehensive availability analysis method for power communication optical cables taking service characteristics into consideration. Background Art

[0002] Power communication optical cables are a crucial carrier medium for power communication services. Effective utilization of optical cable resources is a key means of reducing network operating costs and improving network service quality. The availability of power communication optical cable resources depends not only on the utilization of physical fiber core resources but also on the availability of the cables and the characteristics of the services they carry. To reduce the risk of large-scale service interruptions caused by cable failures, the availability of power communication optical cable resources must comprehensively consider the characteristics of the services carried by the cables, cable reliability, and remaining fiber core resources. If the remaining fiber core resources are used solely as the basis for service deployment, the services carried by the cables are likely to operate at high risk. A power communication optical cable resource availability analysis method that considers service characteristics first calculates and statistically analyzes service volume, cable availability, and the percentage of remaining fiber core resources based on the characteristics of the services currently carried by the cables and the characteristics of the cables themselves. The statistical results are then used to remove extreme data objects, construct an analysis dataset, and standardize the matrix. On this basis, a difference matrix between data objects is constructed, and the optimal number of clusters is estimated using a clustering trend visualization evaluation algorithm. The fuzzy clustering algorithm is executed in multiple cycles to determine the fuzzy set overlap coefficient of fuzzy clustering and obtain the optimized clustering result. Finally, data visualization tools are used to analyze the availability of power communication optical cable resources based on the optimized clustering results.

[0003] At present, there is no method to analyze the effectiveness of power communication optical cable resource utilization by considering the service characteristics of optical cables and using fuzzy clustering technology. Summary of the Invention

[0004] The purpose of the present invention is to provide a comprehensive availability analysis method for power communication optical cables taking into account business characteristics, which helps to accurately identify the business carrying conditions of power communication optical cables and the availability of optical cable resources, reasonably plan and deploy communication services, reduce the risk of optical cable overloading, and improve business reliability, so as to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A comprehensive availability analysis method for power communication optical cables considering service characteristics includes the following steps:

[0007] S1: Extract the characteristics and values of the services carried by the power communication optical cable and calculate the service volume; extract the characteristics and values of the power communication optical cable and calculate the availability of the optical cable; count the utilization of the optical cable core and calculate the remaining proportion of the optical cable core resources;

[0008] S2: Use business volume, optical cable availability, and the remaining ratio of fiber core resources to clean extreme data objects and build an analytical data set;

[0009] S3: Establish a difference matrix and use the clustering trend visualization evaluation algorithm to estimate the number of clusters; loop the fuzzy clustering algorithm to obtain the optimized clustering results;

[0010] S4: Process the clustering results and analyze the availability of power communication optical cable resources in each cluster through data visualization technology.

[0011] Furthermore, the service volume calculated in S1 includes the characteristics of the services carried by the power communication optical cable, including: service level, service type, carrying mode, channel mode, service voltage level and service quantity; among them, the values of service level are: headquarters, branch and provincial company; the values of service type are: relay protection, safety and stability control, dispatching data network, dispatching program control service; the values of carrying mode are: dedicated optical fiber and multiplexing; the values of channel mode are: primary and backup; the values of service voltage level are: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV and AC 220kV;

[0012] Among them, the traffic volume S of power communication optical cable is expressed as:

[0013]

[0014] Where S(i) represents the traffic volume of the i-th optical cable; N(i) is the number of services currently carried by the i-th optical cable; F n is the number of business features involved in the business volume calculation, j=1,2,...,F n , ξ(j) is the weight of the j-th feature, satisfying the conditions: ω(j,k) is the jth feature and the kth value.

[0015] Furthermore, the calculation of optical cable availability in S1 includes the characteristics of power communication optical cables, including: optical cable voltage level, optical cable type, optical cable length and operating life; among them, the values of optical cable voltage level are: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV, AC 220kV, AC 110kV and AC 35kV; the values of optical cable type are: Optical Fiber Composite Overhead Ground Wire (OPGW) and All Dielectric Self-Supporting (ADSS) cable; optical cable length refers to the actual physical length of the optical cable, in kilometers; operating life refers to the number of years from the time of commissioning to the extraction of characteristic data, in years;

[0016] The availability of optical cables is called reliability, and the availability A is expressed as:

[0017] A(i)=1-τ(i)·λ(i)·L(i)

[0018] Where A(i) represents the availability of the i-th optical cable; τ(i) is the average repair time of an optical cable break, which is related to the cable type; L(i) is the length of the optical cable; and λ(i) is the failure rate per km of optical cable, which is related to the cable voltage level, cable type, and years of operation.

[0019] Furthermore, the failure rate per km of optical cable includes:

[0020] The failure rate per km of optical cable is expressed as:

[0021]

[0022] Among them, K V is the voltage level coefficient of the optical cable; is the failure rate of the optical cable in the stable operation stage, that is, the number of interruptions per hour of the optical cable, which is related to the type of optical cable; t is the operating life of the optical cable; t0 is the duration of the stable operation stage of the optical cable, in years; K λ is the annual optical cable aging coefficient, which is related to the optical cable type.

[0023] Furthermore, the remaining proportion of fiber core resources calculated in S1, including the remaining proportion of power communication optical cable core resources, is expressed as:

[0024]

[0025] Among them, η(i) represents the remaining proportion of fiber core resources of the i-th power communication optical cable; n z (i) is the total number of fiber cores in the i-th power communication optical cable; n s(i) is the number of remaining fiber cores in the i-th power communication optical cable, which belongs to the available optical cable resources.

[0026] Furthermore, the specific method in S2 is:

[0027] Through S1, we obtain three vectors: traffic volume S, optical cable availability A, and fiber core remaining resource ratio η. We use mathematical statistics methods to analyze the element values of the three vectors, and regard data outside the 99% confidence interval as extreme data objects and clean them up.

[0028] Assume that the number of optical cables used for analysis after data cleaning is n r , using the three cleaned vectors to form the matrix B', that is Normalize B' to get the matrix The standardized expression is:

[0029]

[0030] matrix This is the constructed analysis data set.

[0031] Furthermore, a difference matrix is established in S3, and the number of clusters is estimated using the cluster trend visualization evaluation algorithm. The specific method is as follows:

[0032] For matrix B, establish the difference matrix in,

[0033]

[0034] The improved Visual Assessment of (Cluster) Tendency (iVAT) algorithm is used for the difference matrix D to obtain the matrix The matrix D * Perform 256-level grayscale transformation to obtain the difference grayscale matrix Grayscale transformation expression is

[0035]

[0036] By observing the difference gray matrix The image is used to estimate the number of clusters C.

[0037] Furthermore, the fuzzy clustering algorithm is executed cyclically in S3 to obtain the optimized clustering results:

[0038] The four parameters of the fuzzy clustering algorithm are the number of clusters C, the fuzzy set overlap coefficient m, and the maximum number of algorithm iterations l. maxand the minimum step distance ε of the objective function, assuming that the value range of the fuzzy set overlap coefficient m is [1.1,5.0], and the interval is divided into N equal intervals m Subinterval, and set the number of fuzzy clustering algorithm cycles to N m ;

[0039] Given the number of clusters C, the fuzzy set overlap coefficient m, and the maximum number of algorithm iterations l max And the minimum step distance ε parameter of the objective function, for the analysis data set matrix Execute the fuzzy clustering algorithm; the objective function is

[0040]

[0041] Among them, b i =(b i,1 ,b i,2 ,b i,3 ) is the i-th vector (row vector) of matrix B, v k is the kth cluster c k The center vector (row vector) of b under the condition of m i Belongs to cluster c k The membership degree, Using μ i,k Construct the membership matrix U p , that is U p ={μ i,k}; Using cluster center vector v k Constitute the central matrix Center, that is

[0042] After successive iterations, the objective function f(n r ,C,m) gradually becomes smaller, the membership matrix U p tends to be stable; when ||U P+1 -U p ||<ε or the number of iterations reaches the maximum value l max When , the iteration stops and the objective function value is considered to be minimum. At this time, the membership matrix and the central matrix is the fuzzy clustering result;

[0043] In order to obtain the optimal clustering results, the algorithm requires N for different m values. m In this cycle, we try the best clustering results under different m values.

[0044] PBMF (Pakhira, Bandy & Maulik-index for Fuzzy C-means Clustering) is selected as the Clustering Validity Index (CVI) of the fuzzy clustering algorithm, and its expression is:

[0045]

[0046] Where C is the number of clusters, determined by the iVAT algorithm; m is the fuzzy set overlap coefficient; n r is the number of rows of matrix B; v is the mean vector of matrix B, that is, When CVI takes the maximum value, the fuzzy clustering result is the best, that is, the optimal clustering result corresponding to m should satisfy the following formula:

[0047] m=arg max CVI(C,m)

[0048] At this time, the membership matrix U and center matrix Center obtained by the fuzzy clustering algorithm are the optimized clustering results.

[0049] Furthermore, the specific method in S4 is as follows:

[0050] According to the optimized fuzzy clustering results, the analysis data set is divided into C clusters using the maximum value of each column element of the membership matrix U. The center of the data point of each cluster is determined by the center matrix Center. The power communication optical cable resource availability analysis can be regarded as a feature interpretation process of these optical cable clusters. With the help of data visualization technology, the optical cable resource availability can be comprehensively analyzed and intuitively displayed. For optical cables with low business volume, high availability, and a high proportion of remaining fiber core resources, the optical cable resource availability can be regarded as high; for optical cables with high business volume, low availability, and a low proportion of remaining fiber core resources, the optical cable resource availability can be regarded as low.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention provides a comprehensive availability analysis method for power communication optical cables that takes service characteristics into consideration. The method extracts the characteristics and values of the services carried by the power communication optical cables, and extracts the characteristics and values of the optical cables themselves, and respectively calculates the service volume, optical cable availability, and the remaining proportion of fiber core resources. The present invention cyclically executes a fuzzy clustering algorithm for the analysis data set constructed for the service volume, optical cable availability, and remaining proportion of fiber core resources, and obtains optimized clustering results. With the help of data visualization technology, the present invention can intuitively and comprehensively display the availability of optical cable resources. By using the present invention, power communication operators can accurately identify the service carrying status of power communication optical cables and the availability of optical cable resources, and can more reasonably plan and deploy communication services, reduce the risk of optical cable overloading, and improve service reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a flow chart of the method of the present invention;

[0054] Figure 2 It is a flow chart of the fuzzy clustering algorithm of the present invention. DETAILED DESCRIPTION

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] See also Figure 1 In an embodiment of the present invention, a method for analyzing the comprehensive availability of power communication optical cables taking into account service characteristics is provided, comprising the following steps:

[0057] Step 1: Extract the characteristics and values of the services carried by the power communication optical cable and calculate the service volume; extract the characteristics and values of the power communication optical cable and calculate the cable availability; count the fiber core utilization of the optical cable and calculate the remaining fiber core resource ratio;

[0058] Step 2: Using business volume, optical cable availability, and the remaining percentage of fiber core resources, we clean extreme data objects and construct an analysis data set.

[0059] Step 3: Establish a difference matrix and use the clustering trend visualization evaluation algorithm to estimate the number of clusters; loop the fuzzy clustering algorithm to obtain the optimized clustering results;

[0060] Step 4: Process the clustering results and analyze the availability of power communication optical cable resources in each cluster through data visualization technology.

[0061] In the above step 1, the service volume is calculated. In the embodiment of the present invention, service level, service type, service bearing mode, service channel mode, service voltage level and service quantity are selected as the characteristics of the power communication optical cable service. Among them, the values of different service characteristics are as follows: the service level values are: headquarters, branch and provincial company; the service type values are: relay protection, safety and stability control, dispatching data network, dispatching program control service, other services, etc.; the service bearing mode values are: dedicated optical fiber and multiplexing; the service channel mode values are: primary and backup; the service voltage level values are: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV and AC 220kV;

[0062] The embodiment of the present invention calculates the service volume based on the above service characteristics and their values. The service volume S of the power communication optical cable is expressed as:

[0063]

[0064] Where S(i) represents the traffic volume of the i-th optical cable; N(i) is the number of services currently carried by the i-th optical cable; F n is the number of business features involved in the business volume calculation, j=1,2,...,F n , the present invention is provided with F n =5; ξ(j) is the weight of the jth feature, satisfying the conditions: ω(j,k) is the jth feature and the kth value. For example, the values of the feature "business level" are headquarters, branches, and provincial companies, and the corresponding ω(j,k) are: 0.9, 0.7, and 0.5 respectively. The values of other business features are assigned in the same way.

[0065] In the calculation of the optical cable availability in step 1 above, the embodiment of the present invention selects the optical cable voltage level, optical cable type, optical cable length, and operating life as the characteristics of the power communication optical cable. The values of different power communication optical cable characteristics are as follows:

[0066] The voltage levels of optical cables are: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV, AC 220kV, AC 110kV, and AC 35kV; the types of optical cables are: Optical Fiber Composite Overhead Ground Wire (OPGW) and All Dielectric Self-Supporting (ADSS); the cable length refers to the actual physical length of the cable, in kilometers; the operating life refers to the number of years from the time the cable was put into operation to the time the characteristic data was extracted, in years.

[0067] Based on the above-mentioned characteristics and values of the power communication optical cable, the present invention calculates the availability of the optical cable. The availability of the optical cable is called reliability. The availability A is expressed as:

[0068] A(i)=1-τ(i)·λ(i)·L(i)

[0069] Where A(i) represents the availability of the i-th optical cable; τ(i) is the average repair time for an optical cable outage, which is related to the cable type and ranges from 8 to 12 hours; λ(i) is the failure rate per km of optical cable, which is related to the cable voltage level, cable type, and years of operation, and is expressed in 1 / hour; L(i) is the length of the optical cable, in km.

[0070] In the above embodiment, the failure rate per km of optical cable includes:

[0071] The failure rate per km of optical cable is expressed as:

[0072]

[0073] Among them, K V K is the voltage level coefficient of the optical cable, which depends on the voltage level of the optical cable. The higher the level, the higher the K. V The smaller; is the failure rate of the optical cable in the stable operation stage, that is, the number of optical cable interruptions per hour, which is related to the optical cable type; the failure rate of OPGW is lower than that of ADSS; t is the operation age of the optical cable; t0 is the duration of the optical cable in the stable operation stage, in years; t0 of OPGW is longer than that of ADSS; K λ The annual aging coefficient of the optical cable is related to the type of optical cable. K is higher for OPGW than for ADSS. λ Be small.

[0074] The remaining proportion of fiber core resources calculated in the above step 1, including the remaining proportion of power communication optical cable core resources, is expressed as:

[0075]

[0076] Among them, η(i) represents the remaining proportion of fiber core resources of the i-th power communication optical cable; n z (i) is the total number of fiber cores in the i-th power communication optical cable; n s (i) is the number of remaining fiber cores in the i-th power communication cable, which is n z (i) The number of available fiber cores after subtracting the number of occupied fiber cores is the available physical optical cable resources.

[0077] The specific method of the above step 2 is:

[0078] The first step is to calculate the three vectors of traffic volume S, optical cable availability A, and fiber core remaining resource ratio η. The three vectors are analyzed using mathematical statistics methods. Data outside the 99% confidence interval is considered as extreme data objects and cleaned.

[0079] Assume that the number of data objects used for analysis after data cleaning is n r , using the three vectors of the cleaned business volume S, the cable availability A and the fiber core remaining resource ratio η, to form the matrix B', that is, Normalize B' to get the matrix The standardized expression is:

[0080]

[0081] matrix This is the constructed analysis data set.

[0082] In step 3 above, a difference matrix is established, and the number of clusters is estimated using the clustering trend visualization evaluation algorithm. The specific method is as follows:

[0083] For matrix B, establish the difference matrix in,

[0084]

[0085] The improved clustering trend visualization evaluation iVAT algorithm is used to reorder the difference matrix D to obtain the reordered matrix For matrix D * Perform 256-level grayscale transformation to obtain the difference grayscale matrix Grayscale transformation expression is

[0086]

[0087] By observing the difference gray matrix The image is used to estimate the number of clusters C.

[0088] In the above step 3, the fuzzy clustering algorithm is executed cyclically to obtain the optimized clustering results:

[0089] The four parameters of the fuzzy clustering algorithm are: the number of clusters C, the fuzzy set overlap coefficient m, the maximum number of algorithm iterations l max and the minimum step distance ε of the objective function; the number of clusters C is estimated by the iVAT algorithm; the maximum number of iterations of the algorithm l max The minimum step distance ε of the objective function is determined by the scale of the actual problem; the value range of the fuzzy set overlap coefficient m is generally [1.1, 5.0]; in order to obtain the best clustering results, the number of fuzzy clustering algorithm cycles is set to N m ; kth m The value of m is k m =1,2,...,N m ;

[0090] Given the number of clusters C, the fuzzy set overlap coefficient m, and the maximum number of algorithm iterations l max And the minimum step distance ε parameter of the objective function, for the analysis data set matrix Execute the fuzzy clustering algorithm; the objective function is

[0091]

[0092] Among them, b i =(b i,1 ,bi,2 ,b i,3 ) is the i-th vector (row vector) of matrix B, v k is the kth cluster c k The center vector (row vector) of b under the condition of m i Belongs to cluster c k The membership degree, Using μ i,k Construct the membership matrix U p , that is U p ={μ i,k}; Using cluster center vector v k Constitute the central matrix Center, that is

[0093] After successive iterations, the objective function f(n r ,C,m) gradually becomes smaller, the membership matrix U p tends to be stable; when ||U P+1 -U p ||<ε or the number of iterations reaches the maximum value l max When , the iteration stops and the objective function value is considered to be minimum. At this time, the membership matrix and the central matrix is the fuzzy clustering result;

[0094] In order to obtain the optimal clustering results, the algorithm requires N for different m values. m In this cycle, we try the best clustering results under different m values.

[0095] PBMF (Pakhira, Bandy & Maulik-index for Fuzzy C-means Clustering) is selected as the Clustering Validity Index (CVI) of the fuzzy clustering algorithm, and its expression is:

[0096]

[0097] Where C is the number of clusters, determined by the iVAT algorithm; m is the fuzzy set overlap coefficient; n r is the number of rows of matrix B; v is the mean vector of matrix B, that is, In N m In the secondary cyclic fuzzy clustering, when CVI takes the maximum value, the fuzzy clustering result is the best, that is, the optimal clustering result corresponding to m should satisfy the following formula:

[0098] m=arg max CVI(C,m)

[0099] At this time, the membership matrix U and center matrix Center obtained by the fuzzy clustering algorithm are the optimized clustering results.

[0100] In step 4 above, the embodiment of the present invention obtains an optimized fuzzy clustering result through a fuzzy clustering process. Based on the maximum value of each column element of the membership matrix U in the clustering result, the algorithm divides the power communication optical cable resource availability into C clusters, with the center matrix Center as the data point center of each cluster. The power communication optical cable resource availability analysis is a feature interpretation process for these data object clusters. With the help of data visualization technology, the availability of optical cable resources can be intuitively and comprehensively displayed. Optical cable clusters with low business volume, high availability, and a high proportion of remaining fiber core resources can be considered to have high optical cable resource availability; optical cable clusters with high business volume, low availability, and a low proportion of remaining fiber core resources can be considered to have low optical cable resource availability. This achieves a classified evaluation of the availability of power communication optical cable resources.

[0101] Design principle: The present invention provides a comprehensive availability analysis method for power communication optical cables that takes into account business characteristics. The basic idea is to combine business and optical cable characteristics to comprehensively solve the problem of rational utilization of power communication optical cable resources. The content includes: extracting the characteristics and values of the business carried by the optical cable and calculating the business volume; extracting the characteristics of the optical cable and their values and calculating the availability of the optical cable; counting the remaining proportion of fiber core resources; using the business volume, optical cable availability and the remaining proportion of fiber core resources to clean extreme data objects and construct an analysis data set; establishing a difference matrix; using a clustering trend visualization evaluation algorithm to estimate the number of clusters; cyclically executing a fuzzy clustering algorithm to obtain optimized clustering results; and analyzing the availability of power communication optical cable resources in each cluster through data visualization technology, which helps to accurately identify the business carrying status of the optical cable and the availability of optical cable resources, reasonably plan and deploy power communication services, and reduce the risk of optical cable load.

[0102] In order to further better explain the present invention, the following specific example analysis and calculation process is provided:

[0103] The embodiment of the present invention selects service level, service type, service carrying mode, service channel mode, service voltage level and service quantity as the characteristics of the power communication optical cable service; the service characteristic weight distribution is as follows: service level ξ(1) = 0.3, service type ξ(2) = 0.3, service carrying mode ξ(3) = 0.05, service channel mode ξ(4) = 0.05, service voltage level ξ(5) = 0.3.

[0104] The values of different service characteristics are as follows:

[0105] (1) Business level: headquarters ω(1,k) = 0.9, branch ω(1,k) = 0.7, provincial company ω(1,k) = 0.5.

[0106] (2) Business type: relay protection ω(2,k) = 0.9, safety and stability control ω(2,k) = 0.8, dispatching data ω(2,k) = 0.7, dispatching program control ω(2,k) = 0.7, other businesses ω(2,k) = 0.5.

[0107] (3) Service carrying mode: dedicated optical fiber ω(3, k) = 0.9, multiplexing ω(3, k) = 0.9.

[0108] (4) Service channel mode: primary ω(4, k) = 0.5, backup ω(4, k) = 0.5.

[0109] (5) Service voltage level: AC 1000kVω(5,k)=0.9, DC ±800kVω(5,k)=0.9, DC ±660kVω(5,k)=0.9, AC 500kVω(5,k)=0.8, AC 220kVω(5,k)=0.7.

[0110] The present invention calculates the service volume based on the above service characteristics and their values. The service volume S of the power communication optical cable is expressed as

[0111]

[0112] Where N(i) is the number of services currently carried by the i-th optical cable.

[0113] The present invention selects the cable voltage level, cable type, cable length and operating life as the characteristics of the power communication optical cable; the values of different power communication optical cable characteristics are as follows:

[0114] (1) Optical cable voltage levels: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV, AC 220kV, AC 110kV and AC 35kV.

[0115] (2) Optical cable types: OPGW and ADSS.

[0116] (3) Cable length: The actual length of the cable at the geographical location, in km.

[0117] (4) Operation life: The number of years from the time the optical cable is put into operation to the time when the characteristic data is extracted, in years.

[0118] The present invention calculates the availability of the optical cable based on the above-mentioned characteristics and values of the power communication optical cable. The optical cable availability A is expressed as

[0119] A(i)=1-τ(i)·λ(i)·L(i)

[0120] Where τ(i) is the average repair time of optical cable interruption, which is related to the type of optical cable and ranges from [8,12] hours.

[0121] For OPGW optical cables, τ(i) = 12 hours; for ADSS optical cables, τ(i) = 8 hours.

[0122] λ(i) is the failure rate per km of optical cable, which is related to the cable voltage level, cable type and operating life, and the unit is "1 / hour"; L(i) is the length of the optical cable, the unit is km.

[0123] The calculation method of λ(i) is:

[0124]

[0125] Among them, K V is the voltage level coefficient of the optical cable, which depends on the voltage level of the optical cable.

[0126] If the voltage level is AC 1000kV, or DC ±800kV, or DC ±660kV, then K V =0.8;

[0127] If the voltage level is AC 500kV, then K V =1.0;

[0128] If the voltage level is AC 220kV or AC 110kV, then K V =1.2;

[0129] If the voltage level is AC 35kV, then K V =1.5.

[0130] It is the failure rate of the optical cable during the stable operation stage and is related to the type of optical cable.

[0131] For OPGW optical cable, For ADSS optical cable,

[0132] t0 is the duration of the stable operation phase of the optical cable, in years.

[0133] For OPGW optical cables, t0 = 12 years; for ADSS optical cables, t0 = 8 years.

[0134] K λ It is the annual aging coefficient of the optical cable and is related to the type of optical cable.

[0135] For OPGW optical cable, K λ =0.1; for ADSS optical cable, K λ =0.5.

[0136] The remaining proportion of power communication optical cable core resources η(i) is expressed as

[0137]

[0138] Among them, n z (i) is the total number of fiber cores in the optical cable; n s (i) is the number of remaining fiber cores in the optical cable.

[0139] Statistical analysis is performed on the three vectors of business volume S, optical cable availability A, and fiber core remaining resource ratio η. Data outside the 99% confidence interval are regarded as extreme data objects and cleaned. If the data object used for analysis after data cleaning is n r Then the three vectors form the matrix B', that is Normalize B' to get the matrix Matrix B is the constructed analysis data set.

[0140] The present invention first establishes a difference matrix for analyzing the data set B Then, the improved clustering trend visualization evaluation iVAT algorithm is used to estimate the optimal number of clusters C; and the number of fuzzy clustering algorithm loop executions N is determined based on the value range of the fuzzy set overlap coefficient m [1.1, 5.0]. m By using the maximum value of the effectiveness index CVI of the PBMF fuzzy clustering algorithm, the optimized clustering result membership matrix U and center matrix Center are obtained.

[0141] The clustering results are processed and the availability of power communication optical cable resources in different clusters is analyzed through data visualization technology.

[0142] See also Figure 2 The flowchart of the fuzzy clustering algorithm in the embodiment of the present invention is as follows: Figure 2 As shown, the following steps are included:

[0143] Step 1: Input the analysis data set B and build the difference matrix of B in, Then, the iVAT algorithm is used to reorder D to obtain the sorted difference matrix In the matrix D * Perform 256-level grayscale transformation to obtain the difference grayscale matrix Finally, display and observe the difference grayscale matrix image, estimate the optimal number of clusters C;

[0144] Step 2: Assume that the number of fuzzy clustering algorithm loop executions is N m , select N at equal intervals on the interval [1.1, 5.0] of the fuzzy set overlap coefficient m msample points, as the m value for each cycle; the kth m The value of m is k m =1,2,...,N m For example, let N m =40, then the m sample point vector V m ={1.1,1.2,...,5.0}; Given the number of clusters C, the fuzzy set overlap coefficient m, the maximum number of iterations and the iteration step as initial parameters, the fuzzy clustering algorithm is executed for a total of N cycles m times; eventually get N m The membership matrix U and cluster center matrix Center under different m values;

[0145] Step 3: The algorithm selects PBMF as the clustering effectiveness index CVI, according to N m The group membership matrix U and cluster center matrix Center are used to calculate the CVI of the fuzzy clustering results for dataset B. The clustering result with the maximum CVI is considered the optimized result. The optimized membership matrix U and cluster center matrix Center serve as the basis for analysis. The algorithm uses data visualization to analyze the availability of optical cable resources.

[0146] Assume that the total number of power communication optical cables analyzed is 732. Considering the characteristics of optical cable service level, service type, service carrying mode, service channel mode, service voltage level and service quantity, the service volume is calculated to obtain the optical cable service volume vector S. Considering the characteristics of optical cable voltage level, optical cable type, optical cable length and operating years, the optical cable availability (reliability) is calculated to obtain the optical cable availability vector A. The remaining proportion of fiber core resources of each optical cable is counted to obtain the remaining proportion vector η of fiber core resources. The three vectors are cleaned according to the 99% confidence interval. The data outside the confidence interval is regarded as extreme data objects and is cleaned. Among the 732 optical cables, the data objects of 4 are cleaned. The number of data objects used for analysis after data cleaning is n. r =728; normalize the matrix B' consisting of the three eigenvectors to obtain the analysis data set matrix

[0147] Establish the difference matrix of B Use the iVAT algorithm to reorder D and reorder the difference matrix D * Perform 256-level grayscale transformation to obtain the difference grayscale matrix Display and observe the image of the difference grayscale matrix G, and estimate the optimal number of clusters C=5.

[0148] Assume that the number of execution cycles of the fuzzy clustering algorithm is N m=40, the fuzzy set overlap coefficient m is evenly spaced from 40 points in the interval [1.1, 5.0] to form the m sample point vector V m ={1.1,1.2,...,5.0}. Assume that the number of clusters C = 5, and the fuzzy set overlap coefficient m∈V m , the maximum number of iterations it max =100, the minimum step distance of the objective function ε=10 -5 The algorithm loops through fuzzy clustering 40 times, generating 40 sets of membership matrices U and cluster center matrices Center. For each of these 40 sets of results, combined with the analysis of the dataset matrix B, the PBMF clustering effectiveness index (CVI) is calculated. The maximum CVI value is 0.073, corresponding to C = 5 and m = 1.2. The membership matrix U and cluster center matrix Center corresponding to C = 5 and m = 1.2 represent the optimized fuzzy clustering results.

[0149] Based on the optimized clustering results, 728 data objects were divided into five clusters. Using data visualization technology, combined with the optical cables and their service carrying conditions, the analysis results of the power communication optical cable resource availability considering service characteristics are as follows:

[0150] Cluster 1 (C1): Contains 188 optical cables, accounting for 25.8%. Its characteristics are low "business volume", high "optical cable availability (reliability)", and high "fiber core resource surplus ratio". This type of optical cable has low business risk and high resource availability.

[0151] Cluster 2 (C2): Contains 64 optical cables, accounting for 8.8%. It is characterized by high "business volume", low "optical cable availability (reliability)", and "fiber core resource remaining ratio" to medium-high. This type of optical cable has high business risk and low resource availability.

[0152] Cluster 3 (C3): Contains 77 optical cables, accounting for 10.6%. Its characteristics are medium "traffic volume", medium "cable availability (reliability)", and low "remaining fiber core resource ratio". This type of optical cable business has medium risk, few spare resources, and medium to low cable resource availability.

[0153] Cluster 4 (C4): Contains 223 optical cables, accounting for 30.6%. It is characterized by low "business volume", high "cable availability (reliability)", and "above-average fiber core resource surplus ratio". This type of optical cable has low business risk, medium-level spare resources, and high cable resource availability.

[0154] Cluster 5 (C5): Contains 176 optical cables, accounting for 24.2%. It is characterized by medium business volume, high optical cable availability (reliability), and a medium-low percentage of remaining fiber core resources. This type of optical cable has medium business risk, low spare resources, and relatively low cable resource availability.

[0155] The above analysis conclusions help power communication operation and maintenance personnel to accurately identify the service carrying status of optical cables, grasp the availability of optical cable resources, plan and deploy power communication services more reasonably, and further reduce the risk of optical cable loading.

[0156] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A comprehensive availability analysis method for power communication optical cables considering service characteristics, characterized in that: The following steps are involved: S1: Extract the characteristics and values of the services carried by the power communication optical cable and calculate the service volume; extract the characteristics and values of the power communication optical cable and calculate the cable availability; Statistics on the utilization of optical cable cores and calculation of the remaining proportion of fiber core resources; The calculated service volume includes the characteristics of the services carried by the power communication optical cable, including: service level, service type, carrying mode, channel mode, service voltage level, and service quantity; the service level is selected as headquarters, branch, and provincial company; the service type is selected as relay protection, safety and stability control, dispatching data network, and dispatching program control service; the carrying mode is selected as dedicated optical fiber and multiplexing; the channel mode is selected as primary and backup; the service voltage level is selected as AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV, and AC 220kV; Among them, the traffic volume S of power communication optical cable is expressed as: Where S(i) represents the traffic volume of the i-th optical cable; N(i) is the number of services currently carried by the i-th optical cable; F n is the number of business features involved in the business volume calculation, j=1,2,...,F n , ξ(j) is the weight of the j-th feature, satisfying the conditions: ω(j,k) is the jth feature and the kth value; The calculation of optical cable availability includes the characteristics of power communication optical cables, including: optical cable voltage level, optical cable type, optical cable length and service life. The values of optical cable voltage level are: AC 1000kV, DC ±800kV, DC ±660kV, AC 500kV, AC 220kV, AC 110kV and AC 35kV; the values of optical cable type are: optical fiber composite overhead ground wire and all-dielectric self-supporting optical cable; optical cable length refers to the actual physical length of the optical cable, in kilometers; service life refers to the number of years from the time of commissioning to the extraction of characteristic data, in years. The availability of optical cables is called reliability, and the availability A is expressed as: A(i)=1-τ(i)·λ(i)·L(i) Where A(i) represents the availability of the i-th optical cable; τ(i) is the average repair time for a cable break, which is related to the cable type; L(i) is the cable length; λ(i) is the failure rate per km of optical cable, which is related to the cable voltage level, cable type, and years of operation. The failure rate per km of optical cable includes: The failure rate per km of optical cable is expressed as: Among them, K V is the voltage level coefficient of the optical cable; is the failure rate of the optical cable in the stable operation stage, that is, the number of interruptions per hour of the optical cable, which is related to the type of optical cable; t is the operating life of the optical cable; t0 is the duration of the stable operation stage of the optical cable, in years; K λ is the annual optical cable aging coefficient, which is related to the type of optical cable; The remaining proportion of fiber core resources, including the remaining proportion of power communication optical cable core resources, is calculated as follows: Wherein, η(i) represents the remaining proportion of fiber core resources of the i-th power communication optical cable; n z (i) is the total number of fiber cores in the i-th power communication optical cable; n s (i) is the number of remaining fiber cores in the i-th power communication optical cable, which belongs to the available optical cable resources; S2: Use business volume, optical cable availability, and the remaining ratio of fiber core resources to clean extreme data objects and build an analytical data set; S3: Establish a difference matrix and use the clustering trend visualization evaluation algorithm to estimate the number of clusters; loop the fuzzy clustering algorithm to obtain the optimized clustering results; S4: Process the clustering results and analyze the availability of power communication optical cable resources in each cluster through data visualization technology.

2. A method for comprehensive availability analysis of power communication optical cables considering service characteristics according to claim 1, characterized in that: The specific method in S2 is: Through S1, we obtain three vectors: traffic volume S, optical cable availability A, and fiber core remaining resource ratio η. We use mathematical statistics methods to analyze the element values of the three vectors, and regard data outside the 99% confidence interval as extreme data objects and clean them up. Assume that the number of optical cables used for analysis after data cleaning is n r , using the three cleaned vectors to form the matrix B', that is Normalize B' to get the matrix The standardized expression is: matrix This is the constructed analysis data set.

3. The method for comprehensive availability analysis of power communication optical cables considering service characteristics according to claim 1, characterized in that: A difference matrix is established in S3, and the number of clusters is estimated using the clustering trend visualization evaluation algorithm. The specific method is as follows: For matrix B, establish the difference matrix in, An improved clustering trend visualization evaluation algorithm is performed on the difference matrix D to obtain the matrix The matrix D * Perform 256-level grayscale transformation to obtain the difference grayscale matrix Grayscale transformation expression is By observing the difference gray matrix The image is used to estimate the number of clusters C.

4. The method for comprehensive availability analysis of power communication optical cables considering service characteristics according to claim 1, characterized in that: The fuzzy clustering algorithm is executed cyclically in S3, and the optimized clustering results are as follows: The four parameters of the fuzzy clustering algorithm are the number of clusters C, the fuzzy set overlap coefficient m, and the maximum number of algorithm iterations l. max and the minimum step distance ε of the objective function, assuming that the value range of the fuzzy set overlap coefficient m is [1.1, 5.0], the interval [1.1, 5.0] is divided into N equal intervals m Subinterval, and set the number of fuzzy clustering algorithm cycles to N m ; Given the number of clusters C, the fuzzy set overlap coefficient m, and the maximum number of algorithm iterations l max And the minimum step distance ε parameter of the objective function, for the analysis data set matrix Execute the fuzzy clustering algorithm; the objective function is Among them, b i =(b i,1 ,b i,2 ,b i,3 ) is the i-th vector of matrix B, v k is the kth cluster c k The center vector of b under the condition of m i Belongs to cluster c k The membership degree, Using μ i,k Construct the membership matrix U p , that is U p ={μ i,k }; Using cluster center vector v k Constitute the central matrix Center, that is After successive iterations, the objective function f(n r ,C,m) gradually becomes smaller, the membership matrix U p tends to be stable; when ||U P+1 -U p ||<ε or the number of iterations reaches the maximum value l max When , the iteration stops and the objective function value is considered to be minimum. At this time, the membership matrix and the central matrix is the fuzzy clustering result; In order to obtain the optimal clustering results, the algorithm requires N for different m values. m In this cycle, we try the best clustering results under different m values. PBMF is selected as the effectiveness indicator of the fuzzy clustering algorithm and its expression is: Where C is the number of clusters, determined by the iVAT algorithm; m is the fuzzy set overlap coefficient; n r is the number of rows of matrix B; v is the mean vector of matrix B, that is, When CVI takes the maximum value, the fuzzy clustering result is the best, that is, the optimal clustering result corresponding to m should satisfy the following formula: m=argmaxCVI(C,m) At this time, the membership matrix U and center matrix Center obtained by the fuzzy clustering algorithm are the optimized clustering results.

5. The method for comprehensive availability analysis of power communication optical cables considering service characteristics according to claim 1, characterized in that: The specific method in S4 is as follows: According to the optimized fuzzy clustering results, the analysis data set is divided into C clusters using the maximum value of each column element of the membership matrix U. The center of the data point of each cluster is determined by the center matrix Center. The power communication optical cable resource availability analysis can be regarded as a feature interpretation process of these optical cable clusters. With the help of data visualization technology, the optical cable resource availability can be comprehensively analyzed and intuitively displayed. For optical cables with low business volume, high availability, and a high proportion of remaining fiber core resources, they can be regarded as having high optical cable resource availability; for optical cables with high business volume, low availability, and a low proportion of remaining fiber core resources, they can be regarded as having low optical cable resource availability.

Images