Method, device and electronic device for calculating equivalent resistance of resistor network

Abstract

This patent application discloses a method for calculating the equivalent resistance of a resistor network, which includes: obtaining the admittance matrix of the resistor network and deleting the rows and columns corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved; determining the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic extraction parameters, and the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors; and calculating the equivalent resistance formed between all target nodes and the reference node based on the eigenvalues ​​and eigenvectors of the admittance matrix to be solved, thereby improving the efficiency of determining the equivalent resistance.

Description

Technical Field

[0001] This patent application belongs to the field of circuit design technology, and in particular relates to a method, device, and electronic device for calculating the equivalent resistance of a resistor network. Background Art

[0002] With the continuous development of advanced semiconductor processes, the power density of large-scale integrated circuits continues to increase, and the width of metal interconnects is becoming narrower and narrower. This has led to a more prominent problem of metal electromigration, which in turn leads to excessive voltage drop between interconnects. Excessive voltage drop can seriously affect chip performance and even cause complete chip failure. To avoid this situation, in circuit design, electromigration and IR drop (EMIR) analysis is performed to ensure the smooth flow of large-scale resistor networks. This analysis determines the equivalent resistance of each node in the large-scale resistor network.

[0003] However, the existing equivalent resistance solution technology has a large amount of calculation and low calculation efficiency. Summary of the Invention

[0004] This patent application provides a method, device, and electronic device for calculating the equivalent resistance of a resistor network to overcome or alleviate the shortcomings of the prior art.

[0005] A method for calculating the equivalent resistance of a resistor network, comprising:

[0006] Obtaining the admittance matrix of the resistor network and deleting the row and column corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved;

[0007] Determining the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic extraction parameters, wherein the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors;

[0008] The equivalent resistance formed between all target nodes and the reference node is calculated according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved.

[0009] Optionally, calculating the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved includes:

[0010] Calculating the values ​​of the diagonal elements in the inverse matrix corresponding to the admittance matrix to be solved according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved;

[0011] The equivalent resistance formed between all target nodes and the reference node is determined according to the values ​​of the diagonal elements.

[0012] Optionally, the method further includes:

[0013] Select some equivalent resistances from all calculated equivalent resistances, perform linear fitting on them based on their precise values, and determine linear fitting parameters;

[0014] Calculate the corrected value of the remaining equivalent resistance based on the linear fitting parameters.

[0015] Optionally, the method further includes:

[0016] generating an equivalent resistance correction value vector according to a partial correction value of the remaining equivalent resistance;

[0017] An error estimate of the equivalent resistance is determined based on the equivalent resistance correction value vector.

[0018] Optionally, the method further includes:

[0019] Calculate the exact value of the equivalent resistance corresponding to the randomly selected target node according to the standard unit vector corresponding to the randomly selected target node and the admittance matrix to be solved;

[0020] Generating an equivalent resistance precise value vector according to the precise value of the equivalent resistance;

[0021] Determining an error estimate of the equivalent resistance based on the equivalent resistance correction value vector includes:

[0022] Determining a norm ratio between the equivalent resistance correction value vector composed of some corrected values ​​of the remaining equivalent resistance and the equivalent resistance precise value vector composed of some precise values ​​of the remaining equivalent resistance;

[0023] An error estimate of the equivalent resistance is determined based on the norm ratio.

[0024] Optionally, the method further includes:

[0025] An adaptive pre-check is performed on the resistor network to preliminarily estimate whether the eigenvalues ​​of the admittance matrix to be solved have a fast decay characteristic. If so, the method is executed.

[0026] A device for calculating equivalent resistance of a resistor network, comprising:

[0027] The first unit is used to obtain the admittance matrix of the resistor network and delete the row and column corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved;

[0028] The second unit is configured to determine the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic obtaining parameters, wherein the characteristic obtaining parameters include a limit on the number of solved eigenvalues ​​and eigenvectors;

[0029] The third unit is configured to calculate the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved.

[0030] An electronic device comprises a memory and a processor, wherein the memory stores a computer executable program, and when the computer executable program is run by the processor, the method according to any one of claims 1 to 6 is executed.

[0031] In the solution provided in the present application, the admittance matrix of the resistor network is obtained, and the rows and columns corresponding to the specified reference node in the admittance matrix are deleted to obtain the admittance matrix to be solved; the eigenvalues ​​and eigenvectors of the admittance matrix to be solved are determined according to the set characteristic extraction parameters, and the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors; based on the eigenvalues ​​and eigenvectors of the admittance matrix to be solved, the equivalent resistance formed between all target nodes and the reference node is calculated, thereby improving the efficiency of determining the equivalent resistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a method for calculating the equivalent resistance of a resistor network according to an embodiment of the present application.

[0033] Figure 2 A schematic diagram of a specific verification scenario provided for this application.

[0034] Figure 3 This is a schematic diagram of a resistor network in an example scenario of an embodiment of the present application.

[0035] Figure 4 This is a schematic diagram of equivalent accuracy in a scenario of Example 1 of the present application.

[0036] Figure 5 This is a schematic diagram of equivalent accuracy in another scenario of an embodiment of the present application. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of this patent application to clearly describe the technical solutions in the embodiments of this patent application. Obviously, the embodiments described are part of the embodiments of this patent application, not all of the embodiments. Based on the embodiments in this patent application, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of this patent application.

[0038] The terms "first," "second," and the like in the specification and claims of this patent application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this patent application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects. For example, the first object can be one or more. In addition, the term "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0039] In the solution provided in the present application, the admittance matrix of the resistor network is obtained, and the rows and columns corresponding to the specified reference node in the admittance matrix are deleted to obtain the admittance matrix to be solved; the eigenvalues ​​and eigenvectors of the admittance matrix to be solved are determined according to the set characteristic extraction parameters, and the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors; based on the eigenvalues ​​and eigenvectors of the admittance matrix to be solved, the equivalent resistance formed between all target nodes and the reference node is calculated, thereby improving the efficiency of determining the equivalent resistance.

[0040] When calculating equivalent resistance, the external power supply method is usually used. In this method, the independent voltage source in the original circuit is short-circuited, the independent current source is disconnected, and the capacitor is disconnected. A current source (or voltage source) is connected between the two ports. The ratio of the voltage and current between the two ports is the equivalent resistance between the two ports. For convenience, the current of the current source is usually 1A. Therefore, the equivalent resistance between port i and port j can be calculated as shown in the following formula:

[0041]

[0042] R ij represents the equivalent resistance between ports i and j, V i and V j are the voltage values ​​at ports i and j respectively.

[0043] In the process of realizing the present application, the inventors found that in circuit analysis, the circuit equation of the node analysis method of the external current source is generally expressed as GV=I, where G is the admittance matrix of the resistor network, the right-hand term I is the independent current source vector, and V is the node voltage vector. In the calculation of the equivalent resistance by the external power source method, the currents at nodes i and j are I respectively. i =1 and I i =-1. Therefore, the node voltage vector can be expressed as:

[0044] V=G -1 I=G -1(e i -e j )

[0045] where e i and e j are the standard unit vectors corresponding to nodes i and j (or they can be called the vector representation of configuring a 1A independent current source to solve the equivalent resistance between nodes i and j). Similarly, for example, there are corresponding standard unit vectors e1 = (1, 0, ..., 0), e2 = (0, 1, ..., 0), ...e n =(0, 0, ..., 1). etc., where n is the total number of nodes; therefore, the equivalent resistance can be expressed as:

[0046] R ij =(e i -e j ) T V=(e i -e j ) T G -1 (e i -e j )

[0047] If we consider a one-to-many situation, such as node j to any other node i, and let node j be grounded, the equivalent resistance calculation of any node i can be simplified to:

[0048]

[0049] in The matrix obtained by deleting the j-th row and j-th column corresponding to node j from the admittance matrix G is the matrix to be solved. for The inverse matrix of , solving the equivalent resistance of any node i to the ground, can be transformed into solving The i-th element on the main diagonal of , by extension, solving the equivalent resistance can be equivalent to solving The values ​​of the diagonal elements of .

[0050] To this end, in the embodiment of the present application, for example, in one possible scenario, if the scenario where the accuracy requirement of equivalent resistance analysis is not high is considered, such as EMIR (Chinese name: electromigration and current resistance drop), since it is required to meet the error of less than 20%, the approximate method proposed in this application can be used to solve The values ​​of the diagonal elements of , while ensuring the accuracy meets the requirements, reduce the amount of data involved in the operation and improve the calculation efficiency.

[0051] Based on the technical essence discovered by the above inventors in the process of realizing this application, the following embodiment of this application provides a solution for calculating the equivalent resistance of a resistor network, which is preferably applied to scenarios where the equivalent resistance is calculated. Of course, it is not limited to scenarios where the accuracy of equivalent resistance analysis is not required to be high. In some scenarios where the accuracy of equivalent resistance analysis is required to be high, the solution of the embodiment of this application can also be used as long as the requirements can be met.

[0052] Figure 1 This is a flow chart of a method for calculating the equivalent resistance of a resistor network according to an embodiment of the present application. Figure 1 As shown, it includes:

[0053] S101, obtaining the admittance matrix of the resistor network, and deleting the row and column corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved;

[0054] S102, determining the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic extraction parameters, wherein the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors;

[0055] S103 : Calculate the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved.

[0056] Optionally, in one embodiment, the calculating the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved includes:

[0057] Calculating the values ​​of the diagonal elements in the inverse matrix corresponding to the admittance matrix to be solved according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved;

[0058] An equivalent resistance formed between the target node and the reference node is determined according to the values ​​of the diagonal elements.

[0059] Optionally, in one embodiment, the method further includes:

[0060] Select some equivalent resistances from all calculated equivalent resistances, perform linear fitting on them based on their precise values, and determine linear fitting parameters;

[0061] Calculate the corrected value of the remaining equivalent resistance based on the linear fitting parameters.

[0062] Optionally, in one embodiment, the method further includes:

[0063] generating an equivalent resistance correction value vector according to a partial correction value of the remaining equivalent resistance;

[0064] An error estimate of the equivalent resistance is calculated based on the equivalent resistance correction value vector.

[0065] Optionally, in one embodiment, the method further includes:

[0066] Calculate the exact value of the equivalent resistance corresponding to the randomly selected target node according to the standard unit vector corresponding to the randomly selected target node and the admittance matrix to be solved;

[0067] An equivalent resistance precise value vector is generated according to the precise value of the equivalent resistance.

[0068] Determining an error estimate of the equivalent resistance based on the equivalent resistance correction value vector includes:

[0069] Determining a norm ratio between the equivalent resistance correction value vector composed of some corrected values ​​of the remaining equivalent resistance and the equivalent resistance precise value vector composed of some precise values ​​of the remaining equivalent resistance;

[0070] An error estimate of the equivalent resistance is determined based on the norm ratio.

[0071] Optionally, in one embodiment, the method further includes:

[0072] Perform an adaptability pre-check on the resistor network to preliminarily estimate whether the eigenvalues ​​of the admittance matrix to be solved have the characteristic of rapid decay. If so, execute the Figure 1 The method shown.

[0073] Applying the above embodiment of the present application to a resistor network with a total number of n nodes, for example, the exemplary principle for calculating the equivalent resistance of each node is briefly as follows:

[0074] Let A be an n-order square moment. If there exists a real number λ and a non-zero vector x such that Ax = λv, then λ is called the eigenvalue of the matrix A and the vector x is the eigenvector of the matrix A. If λ1, λ2...λ n are all eigenvalues ​​of matrix A, v1, v2...v n For all the corresponding eigenvectors, there exists:

[0075] AV=VD

[0076] Where V is an n×n symmetric matrix:

[0077] V=(v1,v2…v n )

[0078] D is an n-order diagonal matrix:

[0079]

[0080] For a symmetric matrix V, there exists: T =V -1 , so we can get:

[0081] A -1 =VD -1 V T

[0082] If D -1 The n eigenvalues ​​in the array decay rapidly from large to small, so we can take the first p eigenvalues ​​λ1....λ p and its corresponding eigenvectors v1, v2, ..., v p Get an A -1 Approximation:

[0083]

[0084] Among them, V p =(v1, v2, ..., v p ),

[0085] Then A -1 The i-th main diagonal element of can be calculated by the following formula:

[0086]

[0087] Among them, e i It is represented as the standard unit vector corresponding to node i (or it is also called the vector representation of configuring a 1A independent current source to solve the equivalent resistance of node i). is the matrix V p The square of the matrix element in row i and column j in D ii is the i-th element on the diagonal of the diagonal matrix D.

[0088] like Then the equivalent resistance of node i is as follows:

[0089]

[0090]

[0091] Therefore, the equivalent resistances of all nodes (i.e., from node 1 to node n) can be composed of a vector R eigs :

[0092]

[0093] In the embodiment of the present application, as described above, since the first p eigenvalues ​​λ1..λ p and its corresponding eigenvectors v1, v2, ..., v pGet an A -1 The premise of the approximation is D -1 The n eigenvalues ​​in the array decay rapidly from large to small. Therefore, correspondingly, in order to verify Whether the eigenvalues ​​of satisfy the following requirement: the arrangement from large to small has the characteristic of rapid decay. A quick verification scheme is proposed as follows:

[0094] Solve once Among them, e1 represents the standard unit vector corresponding to the first node, x1 represents the solution of the above equation, based on Find x1. If the absolute values ​​of the elements in x1 are evenly distributed and there is no property that the elements in some positions are large while the elements in other positions are approximately 0, then it means that The characteristic value of satisfies: the arrangement from large to small has the characteristic of rapid attenuation, then it is determined that the resistor network can adopt the above Figure 1 The method of the embodiment solves the equivalent resistance.

[0095] It should be noted here that, The above verification is explained using the first node as an example. In fact, any node in the resistor network can be selected for the above verification.

[0096] For example, Figure 2 A schematic diagram of a specific verification scenario provided for this application. Figure 2 , in The figure shows the verification case where all element values ​​in x1 are concentrated around -1 and have roughly the same order of magnitude, which means The eigenvalues ​​of can satisfy the following characteristics: the arrangement from large to small has the characteristic of rapid decay.

[0097] Here, it should be noted that the above description is based on an example in which the element values ​​are all concentrated around -1. However, it should be noted that the element values ​​may also be concentrated around other values ​​other than 0.

[0098] Figure 3 This is a schematic diagram of a resistor network in an example scenario of an embodiment of the present application. Figure 3 As shown, it is an m×m random resistor network, where m=500 and all resistors are random numbers between (0, 1]. Let node 1 be the reference node and grounded. Therefore, the equivalent resistance of all remaining nodes relative to the reference node is m 2 -1, and m is a positive integer greater than zero.

[0099] The admittance matrix G of the resistor network is obtained by improving the node analysis method through the circuit analysis method. The rows and columns corresponding to the reference nodes are deleted to obtain the admittance matrix of the resistor network to be solved. at this time The dimension is n×n, where n=m2 -1.

[0100] In this application scenario, in addition to the vector R of the equivalent resistance obtained based on the above embodiment of the present application, eigs Later, due to the In fact, it is an approximate value, so it can be further optimized to improve the accuracy. To this end, the exact value of the equivalent resistance corresponding to some nodes can be calculated R acc On this basis, we further calculate the corresponding correction value R for the corresponding approximate equivalent resistance of other nodes. mod , in order to correct the equivalent resistance corresponding to the other nodes.

[0101] If in an application scenario, based on circuit design experience, the number of eigenvalues ​​and eigenvectors p (i.e. the above-mentioned characteristic parameters) is determined, assuming p = 50, that is, The smallest 50 eigenvalues ​​D 50 and the eigenvector V 50 Based on the above method, Corresponding inverse matrix The values ​​of the diagonal elements of are the vectors forming the equivalent resistance R eigs .

[0102] In getting R eigs , determine the equivalent resistance to be further optimized, such as the equivalent resistance to be accurately solved, for example, the number N of equivalent resistances to be accurately solved c = 20, based on the standard unit vectors corresponding to the 20 equivalent resistance nodes to be determined, or called the 20 right-hand side terms e i (See the above formula, since e i Located on the right side of the formula, so it is referred to as the right side term). On this basis, LDL T Decompose and then perform substitution according to existing technology to accurately calculate N c R acc (Or it can be called the exact value of the equivalent resistance.) Each node to be determined has an exact value.

[0103] After obtaining the accurate value of some equivalent resistance according to the above method, the above R eigs The remaining equivalent resistance in is corrected, that is, from R eigs Select the corresponding Do a linear fit in Indicates the first i1, or i2, ..., or The exact value of the equivalent resistance is calculated by the least square method to obtain a, b (i.e. the above linear fitting parameters). c After the resistance values ​​of the equivalent resistors are fitted, all the equivalent resistors must be corrected to obtain the corrected values ​​of the equivalent resistors. where k = 1, 2, ..., n,

[0104] In order to verify the calculation error estimate err of the embodiment of the present application, it is calculated according to the following formula:

[0105]

[0106] In the above formula, norm represents the norm, where Based on the above application Figure 1 Based on the embodiment method, N is randomly selected r The corrected N r The correction values ​​of the equivalent resistances form a vector (i.e., the equivalent resistance correction value vector), N r A vector of the exact values ​​of equivalent resistance (equivalent resistance exact value vector), The accurate value obtained by the above-mentioned method of calculating the equivalent resistance is required. In order to ensure the calculation efficiency, N r It should not be too large, approximately equal to 10 is enough.

[0107] Based on the description of the embodiment described above, when generating the equivalent resistance correction value vector according to the correction value of the equivalent resistance, all equivalent resistances can be calculated, that is, in addition to the remaining equivalent resistances whose correction values ​​are calculated according to the fitting parameters, the correction value vector can also be calculated according to the LDL. T The equivalent resistance correction value is determined by decomposing the part of the equivalent resistance whose exact value is calculated, and together with the correction value of the remaining equivalent resistance, it forms the equivalent resistance correction value vector R mod Similarly, R acc For all equivalent resistances according to LDL T Decompose the calculation into a vector of exact values, and N is selected additionally to give an estimate of the computational error r A vector consisting of the equivalent resistances calculated for each node.

[0108] Of course, it should be noted here that the above-mentioned method of determining equivalent accuracy based on the norm ratio is only an example and is not a unique limitation.

[0109] Figure 4 This is a schematic diagram of equivalent accuracy in a scenario according to the embodiment of this application. Figure 4As shown, 1000 R acc and R mod The error between the two is 4.9%. The time required for full accurate calculation is 5849.8s. The time required for the method of the embodiment of the application is 17.9s, and the speedup ratio of the algorithm is about 327 times.

[0110] Figure 5 This is a schematic diagram of equivalent accuracy in another scenario (when p=500) of the embodiment of the present application. Figure 5 As shown, R mod With R acc The numerical difference of is smaller, but the time of equivalent resistance calculation increases to 349.14s, and the corresponding speedup ratio decreases to 17.

[0111] See Table 1 below, which summarizes the error and speedup ratio of solving the equivalent resistance based on the present application solution compared with the precise method under different p conditions.

[0112] Table 1 Error and speedup ratio of equivalent resistance calculated based on the proposed solution under different p

[0113] p 10 50 100 500 error 6.2％ 4.8％ 3.9％ 2.1％ Speedup 768 327 168 17

[0114] When p = 50, the test results of calculating the equivalent resistance of random resistor networks of different sizes (expressed by matrix dimensions) using the present application solution are shown in Table 2. As can be seen, as the scale of the resistor network continues to increase, the error of the equivalent resistance calculated using the present application also increases, but it is within 10%, which can basically meet the computational requirements of equivalent resistance analysis in EMIR. At the same time, as shown in Table 2, the speedup ratio increases with the increase in scale, which shows that the present application solution has a better acceleration effect on larger-scale resistor networks.

[0115] Table 2 Error and speedup ratio of equivalent resistance of resistor networks of different sizes using this application solution when p=50

[0116] Matrix dimensions 9,999 249,999 999,999 2,249,999 error 1.3％ 4.8％ 6.9％ 5.9％ Speedup 7 327 1394 3076

[0117] In the above device embodiment, the exemplary explanation of each unit can be found in the above Figure 1-Figure 5 The description of the corresponding embodiment.

[0118] An embodiment of the present application further provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer executable program, and when the computer executable program is run by the processor, it executes any method embodiment of the present application.

[0119] The embodiments of the present patent application are described above in conjunction with the accompanying drawings, but the present patent application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this patent application, ordinary technicians in this field can also make many forms without departing from the purpose of this patent application and the scope of protection of the claims, all of which fall within the scope of protection of this patent application.

Claims

1. A method for calculating the equivalent resistance of a resistor network, characterized in that: include: Get the admittance matrix of the resistor network and delete the rows and columns corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved Determining the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic extraction parameters, wherein the characteristic extraction parameters include a limit on the number of solved eigenvalues ​​and eigenvectors; Calculating the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved, which specifically includes: Let A be an n-order square moment. If there exists a real number λ and a non-zero vector x such that Ax = λv, then λ is called the eigenvalue of the matrix A and the vector x is the eigenvector of the matrix A. If λ1,λ2…λ n are all eigenvalues ​​of matrix A, v1, v2…v n For all the corresponding eigenvectors, there exists: AV=VD Where V is an n×n symmetric matrix: V=(v1,v2…v n ) D is an n-order diagonal matrix: For a symmetric matrix V, there exists: T =V -1 , we can get: A -1 =VD -1 V T If D -1 The n eigenvalues ​​in the array decay rapidly from large to small, and the first p eigenvalues ​​λ1…λ p and its corresponding eigenvectors v1,v2,…,v p Get an A -1 Approximation: Among them, V p =(v1,v2,…,v p ), Then A -1 The i-th main diagonal element of can be calculated by the following formula: Among them, e i It is represented as the standard unit vector corresponding to node i, is the matrix V p The square of the matrix element in row i and column j in D ii is the i-th element on the diagonal of the diagonal matrix D; like Then the equivalent resistance of node i is as follows: All nodes correspond to equivalent resistances forming a vector R eigs :

2. The method according to claim 1, characterized in that The step of calculating the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved includes: Calculating the values ​​of the diagonal elements in the inverse matrix corresponding to the admittance matrix to be solved according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved; The equivalent resistance formed between all target nodes and the reference node is determined according to the values ​​of the diagonal elements.

3. The method according to any one of claims 1-2, characterized in that The method further comprises: Select some equivalent resistances from all calculated equivalent resistances, perform linear fitting on them based on their precise values, and determine linear fitting parameters; Calculate the corrected value of the remaining equivalent resistance based on the linear fitting parameters.

4. The method according to claim 3, characterized in that The method further comprises: generating an equivalent resistance correction value vector according to a partial correction value of the remaining equivalent resistance; An error estimate of the equivalent resistance is determined based on the equivalent resistance correction value vector.

5. The method according to claim 4, characterized in that The method further comprises: Calculate the exact value of the equivalent resistance corresponding to the randomly selected target node according to the standard unit vector corresponding to the randomly selected target node and the admittance matrix to be solved; Generating an equivalent resistance precise value vector according to the precise value of the equivalent resistance; Determining an error estimate of the equivalent resistance based on the equivalent resistance correction value vector includes: Determining a norm ratio between the equivalent resistance correction value vector composed of some corrected values ​​of the remaining equivalent resistance and the equivalent resistance precise value vector composed of some precise values ​​of the remaining equivalent resistance; An error estimate of the equivalent resistance is determined based on the norm ratio.

6. The method according to claim 5, characterized in that The method further comprises: An adaptability pre-check is performed on the resistor network to preliminarily estimate whether the eigenvalues ​​of the admittance matrix to be solved have a fast decay characteristic. If so, the step of obtaining the admittance matrix of the resistor network is performed.

7. A device for calculating equivalent resistance of a resistor network, characterized in that: include: The first unit is used to obtain the admittance matrix of the resistor network and delete the rows and columns corresponding to the specified reference node in the admittance matrix to obtain the admittance matrix to be solved The second unit is configured to determine the eigenvalues ​​and eigenvectors of the admittance matrix to be solved according to set characteristic obtaining parameters, wherein the characteristic obtaining parameters include a limit on the number of solved eigenvalues ​​and eigenvectors; The third unit is configured to calculate the equivalent resistance formed between all target nodes and the reference node according to the eigenvalues ​​and eigenvectors of the admittance matrix to be solved, and includes: Let A be an n-order square moment. If there exists a real number λ and a non-zero vector x such that Ax = λv, then λ is called the eigenvalue of the matrix A and the vector x is the eigenvector of the matrix A. If λ1,λ2…λ n are all eigenvalues ​​of matrix A, v1, v2…v n For all the corresponding eigenvectors, there exists: AV=VD Where V is an n×n symmetric matrix: V=(v1,v2…v n ) D is an n-order diagonal matrix: For a symmetric matrix V, there exists: T =V -1 , we can get: A -1 =VD -1 V T If D -1 The n eigenvalues ​​in the array decay rapidly from large to small, and the first p eigenvalues ​​λ1…λ p and its corresponding eigenvectors v1,v2,…,v p Get an A -1 Approximation: Among them, V p =(v1,v2,…,v p ), Then A -1 The i-th main diagonal element of can be calculated by the following formula: Among them, e i It is represented as the standard unit vector corresponding to node i, is the matrix V p The square of the matrix element in row i and column j in D ii is the i-th element on the diagonal of the diagonal matrix D; like Then the equivalent resistance of node i is as follows: All nodes correspond to equivalent resistances forming a vector R eigs :

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer executable program is stored in the memory, and when the computer executable program is run by the processor, the method according to any one of claims 1 to 6 is executed.

Images