CN102054070A - Method and device for calculating branch current of direct-current working point of nonlinear circuit - Google Patents
Method and device for calculating branch current of direct-current working point of nonlinear circuit Download PDFInfo
- Publication number
- CN102054070A CN102054070A CN2009102096366A CN200910209636A CN102054070A CN 102054070 A CN102054070 A CN 102054070A CN 2009102096366 A CN2009102096366 A CN 2009102096366A CN 200910209636 A CN200910209636 A CN 200910209636A CN 102054070 A CN102054070 A CN 102054070A
- Authority
- CN
- China
- Prior art keywords
- matrix
- branch current
- circuit
- circuit equation
- hand side
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Measurement Of Current Or Voltage (AREA)
- Emergency Protection Circuit Devices (AREA)
Abstract
The invention relates to a method and a device for calculating branch current of a direct-current working point of a nonlinear circuit. The method comprises the following steps of: establishing a first circuit equation matrix according to the nonlinear circuit, and calculating the direct-current working point; storing the calculated and converged state variable; newly increasing a zero volt voltage source at the position to obtain the branch current, and establishing a second circuit equation matrix; mapping the state variable into the second circuit equation matrix; solving a branch current detection matrix and a branch current detection right-hand vector according to the matrix element and the right-hand vector of the second circuit equation matrix; and solving the branch current to be obtained according to the branch current detection matrix and the branch current detection right-hand vector.
Description
Technical field
The invention relates to a kind of computing method and device of non-linear circuit dc point, especially about the branch current computing method and the device of non-linear circuit dc point.
Background technology
(Simulation Program with Integrated Circuit Emphasis HSPICE) before carrying out any form, at first need carry out dc analysis to circuit simulation software, uses the Dc bias point of setting up circuit.With this as the starting point, just can carry out the simulation of other performances such as transient state, ac small signal, noise.The calculating of circuit dc point, its essence are exactly to separate a nonlinear Algebraic Equation set on mathematics.In order to set up the dc analysis point of circuit, HSPICE must find the solution one group of nonlinear equation describing the circuit behavior, it can have by the method for non-linear algebraic equation numerical solution commonly used: Newton iteration method (Newton-Raphson Algorithm, N-R algorithm), continuation method and pseudo-transient state method.Though the ultimate principle of these methods is well-known, how at the characteristics of breadboardin, the algorithm of realizing out having excellent properties and very strong convergence but is the problem that the integrated circuit (IC) design personnel perplex the most always.Particularly, the convergence problem of circuit dc point is a problem the most difficult in the breadboardin.
In the computer simulation of circuit, set up circuit equation with the improvement nodal method usually.The nodal method of analysis is that arbitrary node is a kind of analytical approach of variable independently to the current potential of reference mode in the circuit, if have n node to utilize isolated node row write current (KCL) equation to list (n-1) individual independent equation in the circuit, obtain the voltage and the electric current of each branch road assembly then to obtain the current potential of respective nodes to reference mode.The shortcoming of the nodal method of analysis is that can not directly to handle branch road admittance values such as independent voltage source, null value resistor assembly be infinitely-great assembly, can not directly handle except Voltage-controlled Current Source (Voltage-Controlled-Current-Source, VCCS) controlled source in addition.Therefore, the improvement nodal method of follow-up developments is based on nodal method, and overcome above-mentioned shortcoming, and adopted to list circuit equation by HSPICE.
In the equation that the improvement nodal method is set up, having only the minority components and parts is independent variable (as voltage source, inductance etc.) with the branch current; But in concrete breadboardin process, might need to calculate the electric current of many branch roads.If having voltage source, inductance etc. that branch current is arranged itself on this branch road is the components and parts of independent variable, problem just is easy to solve; If but had only other device, common way would be to insert the voltage source of one zero volt in this branch road.At the branch current of required calculating more after a little while; do not do so and can bring too big problem; but when calculative branch current is quite a lot of; a large amount of zero volt voltage sources that insert bring very big circuit matrix through regular meeting; and cause problems such as the serious decline of matrix computations with precision, Newton iteration can't restrain.
Whether the another kind of method of calculating branch current is in the calculating of each device, no matter need, and all calculates the electric current on each port.After iterative computation convergence, print corresponding current value according to the user is required again.This method need all calculate all branch currents in each iteration.Yet the calculating of branch current over half all is repetition, thereby causes the waste of computational resource and the decline of calculated performance.
Thereby prior art only is to survey branch current by inserting zero volt voltage source, or calculates the electric current of its each port in the calculating of each device, can not obtain the branch current of required non-linear circuit dc point with most economical effective and efficient manner.Therefore, how can reduce operation time and speed of convergence and how to obtain a kind of more efficient methods when HSPICE finds the solution Nonlinear System of Equations, be the problem that industry is paid close attention to always.
Summary of the invention
The objective of the invention is to obtain the branch current of required non-linear circuit dc point with most economical effective and efficient manner.The present invention calculates the branch current that inserts the circuit behind the zero volt voltage source according to the working point of the circuit that does not insert zero volt voltage source, therefore the present invention need not insert the laggard row iteration calculating of zero volt voltage source, also need not be that each port of each device calculates electric current in each iteration.
The branch current computing method of the non-linear circuit dc point of one embodiment of the invention comprise following steps: set up the first circuit equation matrix according to this non-linear circuit, and carry out the calculating of dc point; Store the state variable of calculating after restraining; Newly-increased one zero volts of voltage sources in the position of wishing to get branch current, and set up second circuit equation matrix; This state variable is mirrored second circuit equation matrix; Try to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector; Reach according to this branch current and survey matrix and branch current detection right-hand side vector and try to achieve the branch current of wishing to get.
The branch current calculation element of the non-linear circuit dc point of one embodiment of the invention comprises one first computing unit, a storage element, an assembled unit, one second computing unit and one the 3rd computing unit.This first computing unit is set up the first circuit equation matrix according to this non-linear circuit, and carries out the calculating of dc point.The state variable of this first computing unit of this storage unit stores after calculating convergence.This assembled unit is newly-increased one zero volts of voltage sources in the position of wishing to get branch current, set up second circuit equation matrix, and this state variable is mirrored second circuit equation matrix.This second computing unit tries to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector.The 3rd computing unit surveys matrix according to this branch current and branch current is surveyed the right-hand side vector and tried to achieve the branch current of wishing to get.
Method of the present invention only need be set up twice circuit equation, and owing to need not to insert the zero volt voltage source of surveying branch current in iterative computation, the matrix of circuit becomes very little, can obtain very big alleviating to the requirement of internal memory and matrix computations precision.Because the calculating of branch current is separated from original choosing generation calculating, and the calculating of branch current can not influence the convergence that original circuit calculates.
Description of drawings
Fig. 1 is the branch current calculation flow chart of the non-linear circuit dc point of one embodiment of the invention;
Fig. 2 and Fig. 3 are an object lesson of the present invention; And
Fig. 4 is the branch current calculation element of the non-linear circuit dc point of one embodiment of the invention.
Embodiment
For ease of better understanding spirit of the present invention, it is described further below in conjunction with the preferred embodiments of the present invention.Branch current computing method and device that the present invention is a kind of non-linear circuit dc point in this direction of inquiring into.In order to understand the present invention up hill and dale, will detailed step and composition be proposed in the following description.Obviously, enforcement of the present invention is not defined in the specific details that the technician was familiar with of circuit design.On the other hand, well-known composition or step are not described in the details, with the restriction of avoiding causing the present invention unnecessary.Preferred embodiment meeting of the present invention is described in detail as follows, yet except these were described in detail, the present invention can also implement in other embodiments widely, and scope of the present invention do not limited, and it is as the criterion with claims.
Fig. 1 is the branch current calculation flow chart of the non-linear circuit dc point of one embodiment of the invention.In step 11, set up the first circuit equation matrix according to this non-linear circuit, and carry out the calculating of dc point.Before this, can select to remove the promising zero volt voltage source of finding the solution branch current and inserting, set up circuit equation by improving nodal method, carry out the analog computation of dc point.In step 12, store the state variable of calculating after restraining.This state variable will be saved the number of times of many computings in follow-up calculating.In step 13, newly-increased one zero volts of voltage sources in the position of wishing to get branch current, and can select to set up second circuit equation matrix by improving nodal method.It more than is the preparatory stage of newly-built equation matrix.
In step 14, this state variable is mirrored second circuit equation matrix.Because newly-increased zero volt voltage source does not influence the dc point of circuit, thereby the state variable of being stored can be mirrored newly-built equation matrix.In addition, because insertion is zero volt voltage source, the node voltage value of newly-increased node also can obtain near node.In view of the above, having only the branch current of newly-increased zero volt voltage source in the state variable of novel circuit equation is unknown quantity.
In step 15, try to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector.Because newly-increased zero volt voltage source does not influence the dc point of circuit, the matrix element of this second circuit equation matrix and right-hand side vector will be the function of original state variable.Therefore, for finding the solution the branch current of newly-increased zero volt voltage source, only need from newly-built circuit matrix and right-hand side vector, to isolate the corresponding matrix and the right-hand side vector of finding the solution these newly-increased unknown quantitys, be called branch current and survey matrix and branch current detection right-hand side vector.According to improving nodal method to the value of the filling out rule of independent voltage source as can be known: branch current is surveyed matrix and is constituted a diagonal matrix, wherein diagonal element according to the direction of insertion be respectively+1 or-1; And branch current is surveyed the right-hand side vector can deduct newly-built matrix and previous status variable according to the operation rule of square matrix from newly-built matrix right-hand side vector product and is obtained.
In step 16, survey matrix and branch current detection right-hand side vector and try to achieve the branch current of wishing to get according to this branch current.Because it is a diagonal matrix that branch current is surveyed matrix, it is very easy therefore finding the solution the system of linear equations that is made of branch current detection matrix and branch current detection right-hand side vector.In view of the above, the present invention can be easy to obtain newly-increased branch current.
Fig. 2 and Fig. 3 are an object lesson of the present invention, and wherein Fig. 2 represents original circuit, and Fig. 3 represents the circuit behind the newly-increased zero volt voltage source.In order to simplify processing, Fig. 2 and Fig. 3 do not comprise non-linear element, and wherein n1～n4 represents node, and b1 represents branch road 1.
At first, set up system of equations by improving nodal method according to step 11, as follows:
In step 12, the state variable (V after obtaining restraining via iterative
1V
2V
3V
4I
1)
T., and stored.
In step 13, suppose to find the solution the electric current of branch road 4 and branch road 6, then insert corresponding zero volt voltage source, and set up second circuit equation matrix following (wherein with b4, b6 represents branch road 4 and branch road 6 respectively) by improving nodal method:
In addition, can be worth to increase newly node n5 from the node voltage of adjacent node, the magnitude of voltage V of n6
5, V
6In fact, V
5=V
1, V
6=0.
Because zero volt voltage source does not change the state of circuit, the element vector of the matrix element of newly-built matrix and right-hand side vector will be original state variable (V
1V
2V
3V
4I
1)
T(V
5V
6)
TFunction, with (I
4I
6) irrelevant.Has only (I in the newly-built matrix
4I
6) be unknown number, thus can be with the equation of finding the solution them from wherein separating.Below the transposition arrangement can get:
Method of the present invention only need be set up twice circuit equation, and owing to need not to insert the zero volt voltage source of surveying branch current in iterative computation, the matrix of circuit becomes very little, can obtain very big alleviating to the requirement of internal memory and matrix computations precision.Because the calculating of branch current is separated from original choosing generation calculating, and the calculating of branch current can not influence the convergence that original circuit calculates.
Fig. 4 is the branch current calculation element of the non-linear circuit dc point of one embodiment of the invention, and it comprises one first computing unit 41, a storage element 42, an assembled unit 43, one second computing unit 44 and one the 3rd computing unit 45.
This first computing unit 41 is set up the first circuit equation matrix according to this non-linear circuit, and carries out the calculating of dc point.This first computing unit 41 can select to remove the promising zero volt voltage source of finding the solution branch current and inserting, set up circuit equation by improving nodal method, carry out the analog computation of dc point.This storage element 42 stores the state variable of this first computing unit after calculating convergence.This state variable will be saved the number of times of many computings in follow-up calculating.
This assembled unit 43 is newly-increased one zero volts of voltage sources in the position of wishing to get branch current, can select to set up second circuit equation matrix by improving nodal method, and this state variable is mirrored second circuit equation matrix.Because newly-increased zero volt voltage source does not influence the dc point of circuit, thereby the state variable that this storage element 42 can be stored mirrors newly-built equation matrix.In addition, because insertion is zero volt voltage source, the node voltage value of newly-increased node also can obtain near node.In view of the above, having only the branch current of newly-increased zero volt voltage source in the state variable of novel circuit equation is unknown quantity.
This second computing unit 44 tries to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector.Because newly-increased zero volt voltage source does not influence the dc point of circuit, the matrix element of this second circuit equation matrix and right-hand side vector will be the function of original state variable.Therefore, for finding the solution the branch current of newly-increased zero volt voltage source, only need from newly-built circuit matrix and right-hand side vector, to isolate the corresponding matrix and the right-hand side vector of finding the solution these newly-increased unknown quantitys, be called branch current and survey matrix and branch current detection right-hand side vector.According to improving nodal method to the value of the filling out rule of independent voltage source as can be known: branch current is surveyed matrix and is constituted a diagonal matrix, wherein diagonal element according to the direction of insertion be respectively+1 or-1; And branch current is surveyed the right-hand side vector can deduct newly-built matrix and previous status variable according to the operation rule of square matrix from newly-built matrix right-hand side vector product and is obtained.
The 3rd computing unit 45 surveys matrix according to this branch current and branch current is surveyed the right-hand side vector and tried to achieve the branch current of wishing to get.Because it is a diagonal matrix that branch current is surveyed matrix, it is very easy therefore finding the solution the system of linear equations that is made of branch current detection matrix and branch current detection right-hand side vector.In view of the above, the present invention can be easy to obtain newly-increased branch current.
The present invention obtains the branch current of required non-linear circuit dc point with most economical effective and efficient manner.The present invention calculates the branch current that inserts the circuit behind the zero volt voltage source according to the working point of the circuit that does not insert zero volt voltage source, therefore the present invention need not insert the laggard row iteration calculating of zero volt voltage source, also need not be that each port of each device calculates electric current in each iteration.
Technology contents of the present invention and technical characterstic disclose as above, yet those of ordinary skill in the art still may be based on teaching of the present invention and announcements and done all replacement and modifications that does not deviate from spirit of the present invention.Therefore, protection scope of the present invention should be not limited to the content that embodiment discloses, and should comprise various do not deviate from replacement of the present invention and modifications, and is contained by the present patent application claim.
Claims (16)
1. the branch current computing method of a non-linear circuit dc point is characterized in that comprising following steps:
Set up the first circuit equation matrix according to this non-linear circuit, and carry out the calculating of dc point;
State variable after the convergence that storage calculates;
Newly-increased one zero volts of voltage sources in the position of wishing to get branch current, and set up second circuit equation matrix;
This state variable is mirrored this second circuit equation matrix;
Try to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector; And
Survey matrix and branch current detection right-hand side vector and try to achieve the branch current of wishing to get according to this branch current.
2. method according to claim 1 is characterized in that wherein this branch current detection matrix constitutes a diagonal matrix.
3. method according to claim 2, it is characterized in that the diagonal element of this diagonal matrix wherein is expressed as according to the direction of inserting+1 or-1.
4. method according to claim 1 is characterized in that wherein this branch current is surveyed the right-hand side vector deducts this second circuit equation matrix and state variable for the right-hand side vector of this second circuit equation matrix product.
5. method according to claim 1 is characterized in that wherein iterative computation only comes across the calculating of carrying out dc point.
6. method according to claim 1 is characterized in that wherein not using iterative computation at adding zero volt voltage source in the step of wishing to get branch current to trying to achieve.
7. method according to claim 1 is characterized in that wherein this first circuit equation matrix and second circuit equation matrix are according to improving nodal method and setting up.
8. method according to claim 1 is characterized in that should newly-increased zero node voltage that lies prostrate voltage source being to be obtained by near node wherein.
9. the branch current calculation element of a non-linear circuit dc point is characterized in that comprising:
One first computing unit is set up the first circuit equation matrix according to this non-linear circuit, and is carried out the calculating of dc point;
One storage element stores the state variable after the convergence that this first computing unit calculates;
One assembled unit, newly-increased one zero volts of voltage sources are set up second circuit equation matrix, and this state variable are mirrored this second circuit equation matrix in the position of wishing to get branch current;
One second computing unit tries to achieve according to the matrix element of this second circuit equation matrix and right-hand side vector that a branch current is surveyed matrix and branch current is surveyed the right-hand side vector; And
One the 3rd computing unit is surveyed matrix and branch current detection right-hand side vector and is tried to achieve the branch current of wishing to get according to this branch current.
10. device according to claim 9 is characterized in that wherein this branch current detection matrix constitutes a diagonal matrix.
11. device according to claim 9, it is characterized in that the diagonal element of this diagonal matrix wherein is expressed as according to the direction of inserting+1 or-1.
12. device according to claim 9 is characterized in that wherein this branch current is surveyed the right-hand side vector deducts second circuit equation matrix and state variable for the right-hand side vector of second circuit equation matrix product.
13. device according to claim 9 is characterized in that wherein only this first computing unit carries out iterative computation.
14. device according to claim 9 is characterized in that wherein this second computing unit and the 3rd computing unit do not carry out iterative computation.
15. device according to claim 9 is characterized in that wherein this first circuit equation matrix and second circuit equation matrix are according to improving nodal method and setting up.
16. device according to claim 9 is characterized in that should newly-increased zero node voltage that lies prostrate voltage source being to be obtained by near node wherein.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910209636.6A CN102054070B (en) | 2009-10-30 | 2009-10-30 | Method and device for calculating branch current of direct-current working point of nonlinear circuit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910209636.6A CN102054070B (en) | 2009-10-30 | 2009-10-30 | Method and device for calculating branch current of direct-current working point of nonlinear circuit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102054070A true CN102054070A (en) | 2011-05-11 |
| CN102054070B CN102054070B (en) | 2014-11-26 |
Family
ID=43958390
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910209636.6A Active CN102054070B (en) | 2009-10-30 | 2009-10-30 | Method and device for calculating branch current of direct-current working point of nonlinear circuit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102054070B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104778304A (en) * | 2015-02-27 | 2015-07-15 | 东南大学 | Homotopy method for circuit direct-current analysis of MOS (Metal Oxide Semiconductor) tubes |
| CN106649939A (en) * | 2016-09-28 | 2017-05-10 | 哈尔滨工业大学 | Transmission line iteration-based solving method for 2D axial symmetric nonlinear magnetostatic field model |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7002843B2 (en) * | 2004-01-27 | 2006-02-21 | Sandisk Corporation | Variable current sinking for coarse/fine programming of non-volatile memory |
| CN2938404Y (en) * | 2006-03-16 | 2007-08-22 | 中芯国际集成电路制造(上海)有限公司 | Test device for measuring PN junction current |
-
2009
- 2009-10-30 CN CN200910209636.6A patent/CN102054070B/en active Active
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104778304A (en) * | 2015-02-27 | 2015-07-15 | 东南大学 | Homotopy method for circuit direct-current analysis of MOS (Metal Oxide Semiconductor) tubes |
| CN106649939A (en) * | 2016-09-28 | 2017-05-10 | 哈尔滨工业大学 | Transmission line iteration-based solving method for 2D axial symmetric nonlinear magnetostatic field model |
| CN106649939B (en) * | 2016-09-28 | 2019-11-01 | 哈尔滨工业大学 | The method for solving of 2D Axisymmetric Nonlinear magnetostatic model based on transmission line interation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102054070B (en) | 2014-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Litovski et al. | VLSI circuit simulation and optimization | |
| US7353157B2 (en) | Circuit simulation | |
| Wilczak et al. | C^{r}-Lohner algorithm | |
| Banica et al. | Multiplicity structures on space curves | |
| Padgett et al. | Laws of large numbers for normed linear spaces and certain Fréchet spaces | |
| Derickx et al. | Gonality of the modular curve X1 (N) | |
| Ban et al. | Risk-neutral second best toll pricing | |
| US20090119085A1 (en) | Method and system for modeling dynamic behavior of a transistor | |
| Calvo et al. | A stroboscopic numerical method for highly oscillatory problems | |
| Ortega et al. | Singular reduction of Poisson manifolds | |
| US20100198897A1 (en) | Deriving a function that represents data points | |
| Saleh et al. | Accelerating relaxation algorithms for circuit simulation using waveform-Newton and step-size refinement | |
| CN102054070B (en) | Method and device for calculating branch current of direct-current working point of nonlinear circuit | |
| Meissl | A priori prediction of roundoff error accumulation in the solution of a super-large geodetic normal equation system | |
| Drake et al. | A stable algorithm for divergence-free radial basis functions in the flat limit | |
| Razumchik | Analysis of finite capacity queue with negative customers and bunker for ousted customers using chebyshev and gegenbauer polynomials | |
| Peyrl et al. | Numerical solution of the Hamilton–Jacobi–Bellman equation for stochastic optimal control problems | |
| Lu et al. | Analysis of multiscale integrators for multiple attractors and irreversible Langevin samplers | |
| Bednar et al. | Stability of multi-rate simulation algorithms | |
| US8200470B2 (en) | Systems and methods for current analysis of circuits including relatively large RC networks | |
| Rockmore et al. | Computation of L-series for elliptic curves over function fields. | |
| Zhou et al. | The sandpile group of a family of nearly complete graphs | |
| Hebeker et al. | An adaptive finite element method for unsteady convection-dominated flows with stiff source terms | |
| CN110427644A (en) | A kind of the state equation column write method and system of electric analog system | |
| Ortiz | A variational formulation for convection-diffusion problems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |