A method for optimizing ship power grid topology

Abstract

The present invention provides a method for optimizing the topology of a ship power grid, which includes constructing a ship power grid topology spanning tree, selecting reliable redundant branches to increase the reliability of the ship power grid topology, calculating the reliability of the ship power grid topology after adding reliable redundant branches using the Monte Carlo method, and judging whether the reliability of the ship power grid topology meets the standard. The method effectively solves the problem of the difficulty of optimizing the exhaustive method due to the prominent contradiction between calculation accuracy and calculation complexity, and realizes the topological structure design that takes into account both adaptability and reliability of the whole system network; a relatively complete topological structure optimization design method system is constructed, which provides a relatively complete optimization design technical support for the overall plan of the ship power system.

Description

Technical Field

[0001] The present invention relates to the field of power system analysis and energy management, and in particular to a method for optimizing ship power grid topology. Background Art

[0002] For modern all-electric ships, the network topology of the power system can achieve comprehensive management, flexible configuration and efficient utilization of multiple energy sources. In the military field, it serves as a power supply path connecting ship generators and equipment loads, converting electrical energy to supply propulsion, weapons and daily loads respectively, and ensuring the power supply continuity of important loads through the conversion path; in the civilian field, it accepts different forms of energy provided by traditional diesel engines and new LNG, lithium batteries, etc. for ships, converts these energy sources into electrical energy to supply propulsion and daily loads through inverters, and achieves the energy conservation and emission reduction goals of ships through simplified reconstruction. In addition to the key indicators of network reliability for power supply continuity, it is also necessary to consider the constraints of the small space, limited weight and number of spare parts of the ship. With the increase of equipment redundancy, the total volume and total weight of the whole system will increase simultaneously. As a constraint condition of the optimization problem, it cannot exceed the expected indicators of the ship plan, which is in sharp contradiction with improving power supply continuity, bringing difficulties to the design of network topology. Therefore, it is necessary to develop a new optimization design method to meet the overall needs of ship power systems. However, when studying the optimization design method of the network topology of ship power systems considering reliability and adaptability, there are the following three difficulties:

[0003] 1) The optimization goals of adaptability and reliability are in conflict. If the reliability of the network topology does not meet the standard, the network redundancy needs to be increased, which conflicts with the optimization of the system adaptability and leads to poor adaptability.

[0004] 2) Traditional design methods lack optimization theoretical basis. Traditional design methods generally compare several topological structures manually based on several typical networks. They lack optimization theoretical basis, which makes it difficult for such methods to obtain the optimal topological structure.

[0005] 3) The computational complexity of the exhaustive method increases exponentially. The exhaustive method can design the topology of the multi-source ship distribution network through combinatorial optimization, but the computational complexity increases exponentially, resulting in low optimization efficiency of this type of method.

[0006] At the same time, the significant increase in the scale of modern ship power system networks has made traditional topology design and exhaustive methods no longer applicable. Summary of the invention

[0007] The purpose of the present invention is to provide a ship power grid topology optimization method that takes into account both reliability and adaptability in order to address the defects of the prior art, thereby ensuring the feasibility and optimality of a large-scale network topology structure.

[0008] The present invention provides a method for optimizing ship power grid topology, comprising the following steps:

[0009] 1) Constructing the ship power grid topology spanning tree

[0010] 2) Select reliable redundant branches to increase the reliability of the ship power grid topology in step 1)

[0011] 3) Using the Monte Carlo method to calculate the reliability of the ship power grid topology after adding reliable redundant branches in step 2)

[0012] 4) According to step 3), determine whether the reliability of the ship power grid topology meets the standard.

[0013] Furthermore, the specific process of step 1) is as follows:

[0014] 1.1) Establish the graph theory model G(V,E) of the ship power grid. According to each node {v1,v2,…,v n}Get the node reliability index value {r v1 ,r v2 ,…,r vn} and according to the graph model G(V,E), each branch {e1,e2,…,e m}Get the branch reliability index value {r e1 ,r e2 ,…,r em};

[0015] Where n is the number of nodes; m is the number of branches; V = {v1,v2,…,v n}; E = {e1, e2, …, e m};

[0016] 1.2) If the jth node v j is the i-th branch e i The endpoints satisfy r vj >0 and r ei >0, then judge branch e i It is a reliable redundant branch;

[0017] If r ei =0, then judge branch e i It is an unreliable branch;

[0018] If the jth node v j is the i-th branch e i The endpoints of r vj =0, then judge branch e i It is a non-redundant branch;

[0019] If the above conditions are not met, then judge branch ei It is an unreliable and non-redundant branch;

[0020] Where, j = 1, 2, ..., n, i = 1, 2, ..., m; r vj For node v j Reliability index value; r ei For branch e i Reliability index value of

[0021] 1.3) All reliable redundant branches determined in step 1.2) are recorded as {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}, and record the reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node is {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)*}; m* is the number of reliable redundant branches; n* is the number of associated nodes corresponding to the reliable redundant branches; i*=1,2,…,m*, j*=1,2,…,n*;

[0022] The reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)* The corresponding reliability index value {r e1* ,r e2* ,…,r e(i*)* ,…,r e(m*)*} and reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)* The reliability index value corresponding to {r v1* ,r v2* ,…,r v(j*)* ,…,r v(n*)*} are sorted in descending order, that is, r e1* >r e2* >…>r e(i*)* >…>r e(m*)* 、r v1* >r v2* >…>r v(j*)* >…>r v(n*)*; Select all reliable redundant branches to construct the branch-node association matrix A of the ship power grid;

[0023]

[0024] A is an m*×n* matrix, whose elements are defined as:

[0025]

[0026] 1.4) Let the rank of A be w, perform LU decomposition on A, process one row or one column of A in each step, and complete the decomposition after w steps, obtaining a lower triangular matrix L and an upper triangular matrix U;

[0027]

[0028]

[0029] Among them, L is an m*×w-order matrix, U is a w×n*-order matrix; {e r1* ,e r2* ,…,e rw* ,e r(w+1)* ,…,e r(m*)*} is the reliable redundant branch after LU decomposition, {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)*} is the node associated with the reliable redundant branch after LU decomposition;

[0030] 1.5) Select the reliable redundant branches {e r1* ,e r2* ,…,e rw*}, the node {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)*}The connections constitute the ship power grid topology spanning tree.

[0031] Furthermore, in step 1.4), let z=1, 2, ..., w, then the calculation process of the zth step of LU decomposition includes:

[0032] 1.4.1) Row and column permutations

[0033] For the pivot a in A z,z , check whether there is a p,q≠0(p=z,…,m*;q=z,…,n*); if it exists, swap the zth row and pth row, zth column and qth column of A, and swap the zth row and pth row, zth column and qth column of L; if it does not exist, terminate the decomposition process;

[0034] 1.4.2) Calculate the zth column of L

[0035] By formula l ii,z =a ii,z / a z,z Calculate the element l in L ii,z , where ii = z + 1, ..., m*;

[0036] 1.4.3) Calculate the zth row of U

[0037] The element u in U z,jj Assign the value to the corresponding element a in A z,jj , where jj=z,…,n*;

[0038] 1.4.4) Update elements in A

[0039] The element a in A ii,jj Assign a ii,jj -l ii,z ×a z,jj , where ii = z+1,…,m*, jj = z+1,…,n*;

[0040] Furthermore, the specific process of step 2) is as follows:

[0041] According to the characteristics of the Euler circuit, extract a non-zero element l on the rightmost side of the lower triangular matrix L x,y , select the corresponding branch e rx* It is added to the ship's power grid as a reliable redundant branch, thereby increasing the reliability of the ship's power grid topology; x=w+1,w+2,…,m*, y=1,2,…,w.

[0042] Furthermore, the specific process of step 3) is as follows:

[0043] 3.1) Set the total number of independent repeated experiments k for the Monte Carlo method max and the total number of independent repeated experiments k max The number of successful independent repeated experiments k success , let k = 1 and k success =0, if k max ≥k, then iterate from step 3.2) to step 3.3) and if k max <k, go to step 3.4);

[0044] 3.2) Let k = k + 1. Based on the ship power grid topology, according to the reliability index value corresponding to the node or the reliability index value corresponding to the branch, the available branches are randomly selected, and then the unavailable nodes and the associated branches of the unavailable nodes are randomly removed to form a random network.

[0045] 3.3) Construct the branch-node association matrix A' of the random network, perform the LU decomposition process of step 1.4) on A', and use the upper triangular matrix U to determine whether the source and load nodes belong to the same connected branch, that is, each load has an available power supply path. If so, the independent repeated experiment is successful once, and let k success =k success +1, then return to step 3.2); if not, return directly to step 3.2);

[0046] Among them, the connected branch represents a connected subnetwork;

[0047] 3.4) Divide the number of successful independent repeated experiments by the total number of independent repeated experiments to obtain the available frequency F of the ship power grid topology success , that is, F success =k success / k max .

[0048] Furthermore, in the step 3.2), the probability P=λ / (λ+μ), λ is the fault interval time, and μ is the fault repair time.

[0049] Furthermore, the specific process of step 4) is as follows:

[0050] The ship power grid topology obtained in step 3) is used to characterize the reliability of the ship power grid topology with probability, and is compared with standard indicator data set in advance or provided by traditional experience;

[0051] If the topology reliability meets the standard, the ship power grid topology optimization design is completed and the method process is terminated;

[0052] If the topology reliability does not meet the standard, the ship power grid topology still needs to be optimized, and return to step 2).

[0053] Compared with the prior art, the present invention is based on a ship power grid topology optimization method that takes reliability and adaptability into consideration, which effectively solves the difficulty in finding the optimal solution of the exhaustive method due to the prominent contradiction between calculation accuracy and calculation complexity, and realizes the topology structure design of the entire system network that takes into account both adaptability and reliability, and has automatic optimization capabilities; improves the selection strategy of redundant branches, adopts a matrix calculation method, and avoids the influence of the complicated process of the exhaustive method; constructs a relatively complete topology structure optimization design method system, and provides a relatively complete optimization design technical support for the overall plan of the ship power system; through the effective selection of redundant branches, the difficulty in finding the optimal solution of the exhaustive method due to the prominent contradiction between calculation accuracy and calculation complexity is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the process flow of the ship power grid topology optimization method of the present invention;

[0055] Figure 2 The typical topological structure diagram of a 15-node ship square power distribution network;

[0056] Figure 3 A 15-node ship square power distribution network topology spanning tree constructed for the present invention;

[0057] Figure 4 The optimized topology of the 15-node ship square power distribution network designed for the method of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments to facilitate a clear understanding of the present invention, but they do not constitute a limitation on the present invention.

[0059] like Figure 1 The ship power grid topology optimization method shown includes the following steps:

[0060] 1) Constructing the ship power grid topology spanning tree

[0061] 1.1) Establish the graph theory model G(V,E) of the ship power grid. According to each node {v1,v2,…,v n}Get the node reliability index value {r v1 ,r v2 ,…,r vn} and according to the graph model G(V,E), each branch {e1,e2,…,e m}Get the branch reliability index value {r e1 ,r e2 ,…,r em};

[0062] Where n is the number of nodes; m is the number of branches; V = {v1,v2,…,vn}; E = {e1, e2, …, e m};

[0063] 1.2) If the jth node v j is the i-th branch e i The endpoints satisfy r vj >0 and r ei >0, then judge branch e i It is a reliable redundant branch;

[0064] If r ei =0, then judge branch e i It is an unreliable branch;

[0065] If the jth node v j is the i-th branch e i The endpoints of r vj =0, then judge branch e i It is a non-redundant branch;

[0066] If the above conditions are not met, then judge branch e i It is an unreliable and non-redundant branch;

[0067] Where, j = 1, 2, ..., n, i = 1, 2, ..., m; r vj For node v j Reliability index value; r ei For branch e i Reliability index value of

[0068] 1.3) All reliable redundant branches determined in step 1.2) are recorded as {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}, and record the reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node is {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)*}; m* is the number of reliable redundant branches; n* is the number of associated nodes corresponding to the reliable redundant branches; i*=1,2,…,m*, j*=1,2,…,n*;

[0069] The reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)* The corresponding reliability index value {r e1* ,re2* ,…,r e(i*)* ,…,r e(m*)*} and reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)* The reliability index value corresponding to {r v1* ,r v2* ,…,r v(j*)* ,…,r v(n*)*} are sorted in descending order, that is, r e1* >r e2* >…>r e(i*)* >…>r e(m*)* 、r v1* >r v2* >…>r v(j*)* >…>r v(n*)* ; Select all reliable redundant branches to construct the branch-node association matrix A of the ship power grid;

[0070]

[0071] A is an m*×n* matrix, whose elements are defined as:

[0072]

[0073] 1.4) Let the rank of A be w, perform LU decomposition on A, process one row or one column of A in each step, and complete the decomposition after w steps, obtaining a lower triangular matrix L and an upper triangular matrix U;

[0074]

[0075]

[0076] Among them, L is an m*×w-order matrix, U is a w×n*-order matrix; {e r1* ,e r2* ,…,e rw* ,e r(w+1)* ,…,e r(m*)*} is the reliable redundant branch after LU decomposition, {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)*} is the node associated with the reliable redundant branch after LU decomposition;

[0077] Let z = 1, 2, ..., w, then the calculation process of the zth step of LU decomposition includes:

[0078] 1.4.1) Row and column permutations

[0079] For the pivot a in A z,z , check whether there is a p,q ≠0(p=z,…,m*;q=z,…,n*); if it exists, swap the zth row and pth row, zth column and qth column of A, and swap the zth row and pth row, zth column and qth column of L; if it does not exist, terminate the decomposition process;

[0080] 1.4.2) Calculate the zth column of L

[0081] By formula l ii,z =a ii,z / a z,z Calculate the element l in L ii,z , where ii = z + 1, ..., m*;

[0082] 1.4.3) Calculate the zth row of U

[0083] The element u in U z,jj Assign the value to the corresponding element a in A z,jj , where jj=z,…,n*;

[0084] 1.4.4) Update elements in A

[0085] The element a in A ii,jj Assign a ii,jj -l ii,z ×a z,jj , where ii = z+1,…,m*, jj = z+1,…,n*;

[0086] 1.5) Select the reliable redundant branches {e r1* ,e r2* ,…,e rw*}, the node {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)*}Connect to form a ship power grid topology spanning tree;

[0087] 2) Since the structure of the ship power grid topology spanning tree is radial, the adaptability of the ship power grid topology is optimal but the reliability is insufficient. It is necessary to continue to select redundant branches to increase the reliability. According to the characteristics of the Euler loop, reliable redundant branches are selected in the lower triangular matrix L of the LU decomposition to increase the reliability of the ship power grid topology. The specific method is as follows:

[0088] Extract a non-zero element l from the rightmost side of the lower triangular matrix L x,y , select the corresponding branch e rx* As a reliable redundant branch, it is added to the ship's power grid, thereby increasing the reliability of the ship's power grid topology; x = w + 1, w + 2, ..., m*, y = 1, 2, ..., w;

[0089] According to the characteristics of the Euler circuit, each time a reliable redundant branch is added to the ship power grid topology spanning tree, an Euler circuit is formed, and each branch is traversed only once, so the degree of each node on the Euler circuit is an even number. Correspondingly, in the LU decomposition lower triangular matrix L, the rows corresponding to all branches on the Euler circuit are extracted to form a submatrix, and the sum of the elements in each column is an even number. In general, the closer the non-zero elements in the submatrix are to the right, the more branches may be associated when tracing back the path, and the lower the dependence on the reliability of any branch. Therefore, a non-zero element l corresponding to a reliable redundant branch is extracted on the far right of the lower triangular matrix L. x,y , then by selecting the branch e rx* In general, the single-step improvement in the reliability of the ship power grid topology is the most significant.

[0090] 3) The reliability of the ship power grid topology spanning tree after adding the reliable redundant branch in step 2) is calculated by using the Monte Carlo method, including the following steps:

[0091] 3.1) Set the total number of independent repeated experiments k for the Monte Carlo method max and the total number of independent repeated experiments k max The number of successful independent repeated experiments k success , let k = 1 and k success =0, if k max ≥k, then iterate from step 3.2) to step 3.3) and if k max <k, go to step 3.4);

[0092] 3.2) Let k = k + 1. Based on the ship power grid topology, according to the reliability index value corresponding to the node or the reliability index value corresponding to the branch, the available branches are randomly selected, and then the unavailable nodes and the associated branches of the unavailable nodes are randomly removed to form a random network.

[0093] The availability probability P = λ / (λ+μ), where λ is the fault interval time and μ is the fault repair time;

[0094] 3.3) Construct the branch-node association matrix A' of the random network, perform the LU decomposition process of step 1.4) on A', and use the upper triangular matrix U to determine whether the source and load nodes belong to the same connected branch, that is, each load has an available power supply path. If so, the independent repeated experiment is successful once, and let k success =k success +1, then return to step 3.2); if not, return directly to step 3.2);

[0095] Among them, the connected branch represents a connected subnetwork, that is, there is a path connected by branches between any two nodes;

[0096] 3.4) Divide the number of successful independent repeated experiments by the total number of independent repeated experiments to obtain the available frequency F of the ship power grid topology success , that is, F success =k success / k max According to the law of large numbers, when the total number of independent repeated experiments is large enough, the available frequency F success As an approximation of the probability of availability formed by the reliability of the ship's power grid topology;

[0097] 4) Determining whether the reliability of the designed ship power grid topology meets the standards includes the following steps:

[0098] The ship power grid topology obtained in step 3) is used to characterize the reliability of the ship power grid topology with probability, and is compared with standard indicator data set in advance or provided by traditional experience;

[0099] If the topology reliability meets the standard, the ship power grid topology optimization design is completed and the method process is terminated;

[0100] If the topology reliability does not meet the standard, the ship power grid topology still needs to be optimized, and return to step 2).

[0101] Example system parameters: The ship's square power distribution network contains a total of 15 nodes, arranged in a 5×3 format; the node numbers are arranged from left to right and from top to bottom; each node supplies power to one load; each reliable branch connects adjacent nodes in the horizontal and vertical directions, and there are 22 branches in total; commonly used reliability indicators include the average failure interval time of the ship underway and the average time consumed for repair, and the international community adopts the availability probability (or continuous power supply probability) for unified representation; the availability probability of the 12 nodes associated with the outer ring of the network is set to 98%, and the availability probability of the remaining 3 internal nodes is set to 99%; the availability probability of the 12 reliable branches in the outer ring of the network is set to 98%, and the availability probability of the remaining 10 reliable branches is set to 99%; referring to the data provided by the American Electric Ship Association, the availability probability standard of the overall network topology is set to no less than 95%.

[0102] The specific embodiments of the present invention are given below:

[0103] The optimization design scheme of 15-node ship square power distribution network topology considering reliability and adaptability. Figure 2 Taking the typical topological structure diagram of a 15-node ship square power distribution network as an example, the specific steps to obtain the optimization design scheme are as follows:

[0104] 1) Constructing the ship power grid topology spanning tree

[0105] 1.1) Construct the graph theory model G(V,E) of the ship power grid, V = {v1,v2,…,v 15} and E={e1,e2,…,e 22} represent the node set and branch set in the calibration order, and calculate each node {v1,v2,…,v 15 The corresponding reliability index value {r vc =0.98|c=1,2,3,4,6,7,9,10,12,13,14,15}, {r vd =0.99|d=5,8,11} and each branch {e1,e2,…,e 22 The corresponding reliability index value {r ef =0.98|f=1,2,3,5,7,10,12,15,17,20,21,22}{r eg =0.99|g=4,6,8,9,11,13,14,16,18,19};

[0106] 1.2) For any branch {e i |i=1,…,22} and any node {v j |j=1,…,15}, if node v j It is branch e i The endpoint of r vj >0, r ei >0 holds;

[0107] To simplify the example, the unreliable or non-redundant branches in the initial network topology have been removed. i All of them are reliable redundant branches;

[0108] 1.3) Associate the nodes {v5,v8,v 11}arranged in nodes {v1,v2,v3,v4,v6,v7,v9,v 10 ,v 12 ,v 13 ,v 14 ,v 15}, change the branch {e4,e6,e8,e9,e11 ,e 13 ,e 14 ,e 16 ,e 18 ,e 19}arranged in the branch {e1,e2,e3,e5,e7,e 10 ,e 12 ,e 15 ,e 17 ,e 20 ,e 21 ,e 22} before, the reliability indexes of nodes and branches are arranged in descending order; all branches are selected to construct the branch-node association matrix A of the ship power grid;

[0109] Therefore, the correlation matrix A is represented as a 22×15 matrix, and its elements are composed as follows:

[0110]

[0111] 1.4) Perform LU decomposition of matrix A. After 14 steps, the result is output as a 22×14 unit lower triangular matrix L and a 14×15 upper triangular matrix U, whose elements are as follows:

[0112]

[0113]

[0114] 1.5) Select the branches {e1,e3,e 17 ,e 20 ,e 22 ,e 21 ,e 15 ,e 16 ,e 12 ,e 10 ,e 11 ,e7,e2,e8}, connect the 15 nodes to form the spanning tree of the ship power grid, such as Figure 3 As shown, the topology is radial, and the overall network availability probability is the product of the availability probabilities of all nodes and branches corresponding to the equipment, calculated as (98%) 12 ×(99%) 3 ×(98%) 11 ×(99%) 3 =59.2%, it is necessary to continue to select reliable redundant branches to increase reliability;

[0115] 2) Extract a non-zero element l from the rightmost side of the triangular matrix L under LU decomposition 20,14, select the corresponding branch e6 as a reliable redundant branch to join the ship power grid, thereby increasing the reliability of the topology;

[0116] 3) Reliability calculation steps

[0117] 3.1) Set the number of independent repeated experiments k for the Monte Carlo method max =10 6 and the number of independent repeated experiments k max The number of successes k success , let k = 1 and k success =0, if k≤k max , then iterate the following steps 3.2) to 3.3) in a loop; if k>k max , then go to step 3.4);

[0118] 3.2) Let k = k + 1. Based on the network topology, according to the reliability index value corresponding to the node or the reliability index value corresponding to the branch, first randomly select the available branches, and then randomly remove the unavailable nodes and the associated branches of the unavailable nodes to form a random network;

[0119] 3.3) Construct the branch-node association matrix A' of the random network, perform LU decomposition on A', and use the upper triangular matrix U to determine whether all nodes belong to the same connected branch. If so, let k success =k success +1, then return to step 3.1); if not, return directly to step 3.2);

[0120] For example, the elements of the upper triangular matrix U obtained from an independent repeated experiment are as follows:

[0121]

[0122] It can be seen that all nodes can be connected only by merging the last column of non-zero elements of U corresponding to node v4, so all nodes belong to the same connected branch. Let k success =k success +1;

[0123] On the contrary, if the independent repeated experiment makes node v4 an unavailable node, then node v5 and other nodes belong to different connected branches, k success remain unchanged;

[0124] 3.4) The number of successful independent repeated experiments k success = 81604 divided by the total number of independent repeated experiments k max =10 6 , get the available frequency F of the ship power grid success =k success / kmax = 81.6%, as an approximation of the probability of network availability;

[0125] 4) Reliability compliance assessment steps

[0126] 4.1) Using step 3) to obtain the network availability probability F success =81.6% characterizes the topology reliability and compares it with the pre-set reliability standard value of 95%;

[0127] 4.2) The topology reliability does not meet the standard, so the current network topology still needs to be optimized, so return to step 2).

[0128] Through continuous loop iteration, the non-zero elements l are continuously extracted from the rightmost side of the triangular matrix L under the initial LU decomposition. 21,14 , l 22,14 , l 15,13 , l 18,11 , select the corresponding branches e9, e4, e5, e 13 Join the ship's power grid as a reliable redundant branch to increase the reliability of the topology;

[0129] By executing step 3), the probability of ship power grid availability is 95.1%. According to step 4), the topology reliability has reached the standard. Therefore, the ship power grid topology optimization design is completed. The network topology is as follows: Figure 4 As shown, the method process terminates;

[0130] Figure 3 and Figure 4 The 15-node ship square distribution network topology structure with optimal adaptability and reliability and its optimization design results are given respectively; from the comparison results of reliability indicators, it can be seen that the above-proposed optimization design method can effectively improve the reliability of the network topology structure under the condition of ship power system adaptability.

Claims

1. A method for optimizing ship power grid topology, characterized in that it comprises the following steps: 1) Constructing the ship power grid topology spanning tree 2) Select reliable redundant branches to increase the reliability of the ship power grid topology in step 1) 3) Using the Monte Carlo method to calculate the reliability of the ship power grid topology after adding reliable redundant branches in step 2) 4) According to step 3), determine whether the reliability of the ship power grid topology meets the standard; The specific process of step 1) is as follows: 1.1) Establish the graph theory model G(V,E) of the ship power grid. According to each node {v 1 ,v 2 ,…,v n}Get the node reliability index value {r v1 ,r v2 ,…,r vn} and according to the graph model G(V,E) each branch {e 1 ,e 2 ,…,e m}Get the branch reliability index value {r e1 ,r e2 ,…,r em}; Where n is the number of nodes; m is the number of branches; V = {v 1 ,v 2 ,…,v n}; E = {e 1 ,e 2 ,…,e m}; 1.2) If the jth node v j is the i-th branch e i The endpoints satisfy r vj >0 and r ei >0, then judge branch e i It is a reliable redundant branch; If r ei =0, then judge branch e i It is an unreliable branch; If the jth node v j is the i-th branch e i The endpoints of r vj =0, then judge branch e i It is a non-redundant branch; If the above conditions are not met, then judge branch e i It is an unreliable and non-redundant branch; Where, j = 1, 2, ..., n, i = 1, 2, ..., m; rvj For node v j Reliability index value; r ei For branch e i Reliability index value of 1.3) All reliable redundant branches determined in step 1.2) are recorded as {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}, and record the reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node is {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)*}; m* is the number of reliable redundant branches; n* is the number of associated nodes corresponding to the reliable redundant branches; i*=1,2,…,m*, j*=1,2,…,n*; The reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}corresponding reliability index value {r e1* ,r e2* ,…,r e(i*)* ,…,r e(m*)*} and reliable redundant branch {e 1* ,e 2* ,…,e (i*)* ,…,e (m*)*}The corresponding associated node {v 1* ,v 2* ,…,v (j*)* ,…,v (n*)* The reliability index value corresponding to {r v1* ,r v2* ,…,r v(j*)* ,…,r v(n*)*} are sorted in descending order, that is, r e1* >r e2* >…>r e(i*)* >…>r e(m*)* 、r v1* >r v2* >…>r v(j*)* >…>r v(n*)* ; Select all reliable redundant branches to construct the branch-node association matrix A of the ship power grid; A is an m*×n* matrix, whose elements are defined as: 1.4) Let the rank of A be w, perform LU decomposition on A, process one row or one column of A in each step, and complete the decomposition after w steps, obtaining a lower triangular matrix L and an upper triangular matrix U; in, L is an m*×w matrix, and U is a w×n* matrix; {e r1* ,e r2* ,…,e rw* ,e r(w+1)* ,…,e r(m*)* } is the reliable redundant branch after LU decomposition, {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)* } is the node associated with the reliable redundant branch after LU decomposition; 1.5) Select the reliable redundant branches {e r1* ,e r2* ,…,e rw* }, the node {v r1* ,v r2* ,…,v rw* ,v r(w+1)* ,…,v r(n*)* }The connections constitute the ship power grid topology spanning tree.

2. According to the ship power grid topology optimization method of claim 1, Features: In step 1.4), let z = 1, 2, ..., w, then the calculation process of the zth step of LU decomposition is include: 1.4.1) Row and column permutations For the pivot a in A z,z , check whether there is a p,q ≠0, p=z,…,m*; q=z,…,n*; if it exists, swap the zth row and pth row, zth column and qth column of A, and swap the zth row and pth row, zth column and qth column of L; if it does not exist, terminate the decomposition process; 1.4.2) Calculate the zth column of L By formula l ii,z =a ii,z / a z,z Calculate the element l in L ii,z , where ii = z + 1, ..., m*; 1.4.3) Calculate the zth row of U The element u in U z,jj Assign the value to the corresponding element a in A z,jj , where jj=z,…,n*; 1.4.4) Update elements in A The element a in A ii,jj Assign a ii,jj -l ii,z ×a z,jj , where ii=z+1,…,m*, jj=z+1,…,n*.

3. According to the ship power grid topology optimization method of claim 1, Features: The specific process of step 2) is as follows: According to the characteristics of the Euler circuit, extract a non-zero element l on the rightmost side of the lower triangular matrix L x,y , select the corresponding branch e rx* It is added to the ship's power grid as a reliable redundant branch, thereby increasing the reliability of the ship's power grid topology; x=w+1,w+2,…,m*, y=1,2,…,w.

4. According to the ship power grid topology optimization method of claim 3, Features: The specific process of step 3) is as follows: 3.1) Set the total number of independent repeated experiments k for the Monte Carlo method max and the total number of independent repeated experiments k max The number of successful independent repeated experiments k success , let k = 1 and k success =0, if k max ≥k, then iterate from step 3.2) to step 3.3) and if k max <k, go to step 3.4); 3.2) Let k = k + 1. Based on the ship power grid topology, according to the reliability index value corresponding to the node or the reliability index value corresponding to the branch, the available branches are randomly selected, and then the unavailable nodes and the associated branches of the unavailable nodes are randomly removed to form a random network. 3.3) Construct the branch-node association matrix A' of the random network, perform the LU decomposition process of step 1.4) on A', and use the upper triangular matrix U to determine whether the source and load nodes belong to the same connected branch, that is, each load has an available power supply path. If so, the independent repeated experiment is counted as successful once, and let k success =k success +1, then return to step 3.2); if not, return directly to step 3.2); Among them, the connected branch represents a connected subnetwork; 3.4) Divide the number of successful independent repeated experiments by the total number of independent repeated experiments to obtain the available frequency F of the ship power grid topology success , that is, F success =k success / k max .

5. According to the ship power grid topology optimization method of claim 4, Features: In the step 3.2), the probability P=λ / (λ+μ), λ is the fault interval time, and μ is the fault repair time.

6. According to the ship power grid topology optimization method of claim 4, Features: The specific process of step 4) is as follows: The ship power grid topology obtained in step 3) is used to characterize the reliability of the ship power grid topology with probability, and is compared with standard indicator data set in advance or provided by traditional experience; If the topology reliability meets the standard, the ship power grid topology optimization design is completed and the method process is terminated; If the topology reliability does not meet the standard, the ship power grid topology still needs to be optimized, and return to step 2).

Images