Fast model generating and solving method for security-constrained power system operation simulation

• US 10,289,765 B2
• Filed: 08/26/2016
• Issued: 05/14/2019
• Est. Priority Date: 09/23/2015
• Status: Active Grant

• Current Assignee: Tsinghua University
• Original Assignee: Tsinghua University
• Inventors: Zhang, Ning, Kang, Chongqing, Xia, Qing, Du, Ershun
• Primary Examiner(s): Wathen, Brian W
• Application Number:
  US15/248,019
• Publication Number:
  US 20170083648A1
• Time in Patent Office: 991 Days
• CPC Class Codes: G06F 17/11 for solving equations , e.g. nonlinear equations, general mathematical optimization problems optimization specially adapted for a specific administrative, business or logistic context G06Q10/04
• CPC Class Codes: G06F 2119/06 Power analysis or power optimisation
• CPC Class Codes: G06F 30/367 Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
• CPC Class Codes: H02J 2203/20 Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
• CPC Class Codes: H02J 3/007 Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources for providing uninterruptable power supply H02J9/00
• CPC Class Codes: Y02E 60/00 Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
• CPC Class Codes: Y04S 40/20 Information technology specific aspects, e.g. CAD, simulation, modelling, system security

Abstract

The present disclosure provides a fast model generating and solving method for security-constrained power system operation simulation, which includes: obtaining information of all branches and nodes which are involved during operation simulation time period, calculating original-node impedance matrix, load shifting distribution factor original matrix and generator shifting distribution factor original matrix of all involved branches; correcting the load shifting distribution factor original matrix and the generator shifting distribution factor original matrix according to the on-off state of branches; obtaining output of each generator unit at each time period according to no-security-constraint unit commitment model, and determining overload of each branch again; solving iteratively until no branch is overloaded, and obtaining output of each generator unit at each time period under security constraint of operation simulation for the current simulation day, performing operating simulation for the rest simulation days to obtain security-constraint operation simulation result for the whole year.

Claims

1. A method for a security-constrained power system operation simulation of a security-constrained power system comprising a plurality of generator units, branches, and nodes, the method comprising the following steps:

1) receiving information from the branches and the nodes during an operation simulation time period, storing the information on a non-transitory computer-readable storage medium, and calculating, with one or more processors, an original-node impedance matrix, load shifting distribution factor original matrix, and generator shifting distribution factor original matrix of the branches according to connections between the branches and the nodes and the reactance of each branch;

2) starting, with the one or more processors, a day-by-day operation simulation calculation by;

obtaining on-off state of branches in a current simulation day;

correcting the load shifting distribution factor original matrix and the generator shifting distribution factor original matrix according to the on-off state of branches to obtain load shifting distribution factor matrix and generator shifting distribution factor matrix in the current simulation day after considering the on-off state of branches;

3) obtaining, with the one or more processors, an output of each generator unit at each time period by using a mixed integer programming algorithm according to no-security-constraint unit commitment model;

calculating, with the one or more processors, each branch power flow for each time period according to the output of each generator unit at each time period and node loads;

determining, with the one or more processors, an overload of each branch according to the branch power flow and branch power flow limits;

generating, with the one or more processors, a security constraint in the unit commitment model according to rows of the load shifting distribution factor matrix and the generator shifting distribution factor matrix corresponding to the overloaded branches;

obtaining, with the one or more processors, the output of each generator unit at each time period by using the mixed integer programming algorithm according to the no-security-constraint unit commitment model again;

determining, with the one or more processors, the overload of each branch again;

solving, with the one or more processors, the output of each generator unit at each time period iteratively until no branch is overloaded;

obtaining, with the one or more processors, the output of each generator unit at each time period under security constraint for completion of operation simulation for the current simulation day;

performing, with the one or more processors, an operating simulation for the remaining simulation days to obtain security-constraint operation simulation results for a whole year.

2. The method of claim 1, wherein the step 1) comprises the following steps:

1.1) generating, with the one or more processors, a branch admittance matrix according to the reactance of each branch, as following;

3. The method of claim 2, wherein the step 2) comprises the following steps:

defining Ω

;

{b₁, b₂, . . . , b_B} as a set of disconnected branches in the current simulation day, and b₁ as a current disconnected branch in the power system, and x_b1 as impedance of b₁, and i as a beginning node of b₁, and j as an end node of b₁;

2.2) calculating, with the one or more processors, an intermediate variable x_b1′

of b₁, as following;

x_b1′

=−

x_b1+X_ii+X_jj−

X_ij−

X_ji;

where, in the above expression, X_ii, X_jj, X_ij, and X_ji are elements corresponding to the original-node impedance matrix X;

2.3) calculating, with the one or more processors, a correction matrix X′

of node impedance matrix considering the disconnected branch b₁, and calculating a node impedance matrix X_b1 considering the disconnected branch b₁ according to the correction matrix X′ and

the original-node impedance matrix X, as following;

4. The method of claim 3, wherein the step 3) comprises the following steps:

3.1) defining a security constraint set Θ

of the power system as being equal to Ø

, Ø

is null set, and defining a calculation loop variable s as being equal to zero, and calculating, with the one or more processors, the no-security-constraint unit commitment model F≤

G_GP^T−

G_DL^T≤

F to obtain output X⁽⁰⁾ of each generator unit, and calculating, with the one or more processors, the branch power flow according to X⁽⁰⁾ and L, as following;

F⁽⁰⁾=G_G,bBp^(0)T−

G_D,bBL^T;

determining, with the one or more processors, an overload of each branch according to F⁽⁰⁾, F and F, and recording set of overloaded branches as Θ

⁽¹⁾;

if Θ

⁽¹⁾ is null set, going to step 3.3), and if Θ

⁽¹⁾ is not null set, updating the security constraint set Θ

=Θ

⁽¹⁾;

defining g_G,bB,l as the elements at l-th row in G_G,bB, and defining g_D,bB,l as the elements at the l-th row in G_D,bB, and defining f_l as the l-th elements of F, and defining f_l as the l-th elements of F, and calculating, with the one or more processors, a security-constrained unit commitment model for the d-th day, as following;

f_l≤

g_G,bB,lP^(0)T−

g_D,bB,lL^T≤

f_l;

solving, with the one or more processors, the unit commitment model considering Θ

to obtain optimal solution P⁽¹⁾ of the unit commitment model considering Θ

in the first iteration;

3.2) defining s=s+1 to enter the next iteration;

for the s-th iteration, calculating the branch power flow according to P^(s) and L, as following;

F^(s)=G_G,bBP^(s)T−

G_D,bBL^T;

determining, with the one or more processors, the overload of each branch according to F^(s), F, and F, and recording the set of overloaded branches as Θ

^(s);

if Θ

^(s) is null set, going to the step 3.3), and if Θ

^(s) is not null set, updating the security constraint set Θ

=Θ

<sup>(s−

1)</sup>, and calculating the security-constrained unit commitment model for the d-th day, with the one or more processors, as following;

f_l≤

g_G,bB,lP^(s)T−

g_D,bB,lL^T≤

f_l,l∈

Θ

^(s);

solving, with the one or more processors, the unit commitment model considering Θ

to obtain P^(s+1), and repeating the step 3.2);

3.3) considering P^(s) as the optimal solution of the security-constrained unit commitment model, and defining d=d+1 to go to the step 2.1) to calculate, with the one or more processors, the operation simulation for the (d+1)-th day.

5. The method of claim 1, wherein the branch is defined by transmission lines, cables, transformers, and power transmission equipment connected to two buses, and the node is defined by one bus in the security-constrained power system.

6. A non-transitory computer-readable storage medium, having stored therein instructions that, when executed by a processor of a device, causes the device to perform a fast model generating and solving method for a security-constrained power system operation simulation;

the security-constrained power system comprising a plurality of generator units, branches, and nodes;

wherein the fast model generating and solving method comprises the following steps;

1) obtaining information of all branches and nodes which are involved during an operation simulation time period, calculating an original-node impedance matrix, load shifting distribution factor original matrix, and generator shifting distribution factor original matrix of all involved branches according to connections between the branches and the nodes and reactance of each branch;

2) starting a day-by-day operation simulation calculation by;

obtaining an on-off state of branches in a current simulation day, correcting the load shifting distribution factor original matrix and the generator shifting distribution factor original matrix according to the on-off state of branches to obtain load shifting distribution factor matrix and generator shifting distribution factor matrix in the current simulation day after considering the on-off state of branches;

3) obtaining an output of each generator unit at each time period by using a mixed integer programming algorithm according to no-security-constraint unit commitment model;

for each time period, calculating each branch power flow according to the output of each generator unit at each time period and node loads;

determining an overload of each branch according to the branch power flow and branch power flow limits;

generating a security constraint in the unit commitment model according to rows of the load shifting distribution factor matrix and the generator shifting distribution factor matrix corresponding to the overloaded branches;

obtaining the output of each generator unit at each time period by using the mixed integer programming algorithm according to the no-security-constraint unit commitment model again; and

then, determining the overload of each branch again;

solving the output of each generator unit at each time period iteratively until no branch is overloaded, and obtaining the output of each generator unit at each time period under security constraint for completion of operation simulation for the current simulation day;

performing operating simulation for remaining simulation days to obtain security-constraint operation simulation results for a whole year.

7. The non-transitory computer-readable storage medium of claim 6, wherein the step 1) comprises the following steps:

1.1) generating branch admittance matrix y according to the reactance of each branch;

1.2) generating node-branch incidence matrix and generator-node incidence matrix according to the connections between the branches and the nodes;

wherein the step 1.2) comprises the following step;

generating branch-node incidence matrix H_l for each branch;

generating the node-branch incidence matrix A by using the branch-node incidence matrix H_l;

setting the node connected to the generator unit with maximum capacity as a relaxation node î and

obtaining reduced-order node-branch incidence matrix Ã

according to the relaxation node î and

the node-branch incidence matrix A;

establishing generator-node incidence matrix R_m for each generator and generating the generator-node incidence matrix A_G by using the generator-node incidence matrix R_m;

obtaining reduced-order generator-node incidence matrix Ã

_G according to the generator-node incidence matrix A_G and the relaxation node î

;

1.3) obtaining the generator shifting distribution factor original matrix G_G and the load shifting distribution factor original matrix G_D according to the branch admittance matrix y and the node-branch incidence matrix A;

wherein the step 1.3) comprises the following steps;

calculating the original-node impedance matrix X, load shifting distribution factor original reduced-order matrix {tilde over (G)}_D, and generator shifting distribution factor original reduced-order matrix {tilde over (G)}_G;

obtaining the load shifting distribution factor original matrix G_D according to the load shifting distribution factor original reduced-order matrix {tilde over (G)}_D and the relaxation node î and

obtaining the generator shifting distribution factor original matrix G_G according to the generator shifting distribution factor original reduced-order matrix {tilde over (G)}_G and the relaxation node î

.

8. The non-transitory computer-readable storage medium of claim 6, wherein the step 2) comprises the following steps:

2.1) defining Ω

;

{b₁, b₂, . . . b_B} as a set of disconnected branches in the current simulation day, and b₁ as a current disconnected branch in the power system, and x_b1 as impedance of b₁, and i as a beginning node of b₁, and j as an end node of b₁;

2.2) calculating an intermediate variable x_b1′

of b₁;

2.3) calculating a correction matrix X′

of node impedance matrix considering the disconnected branch b₁, and calculating a node impedance matrix X_b1 considering the disconnected branch b₁ according to the correction matrix X′ and

the original-node impedance matrix X;

2.4) calculating correction matrix {tilde over (G)}_D′

of load shifting distribution factor considering the disconnected branch b₁ according to the load shifting distribution factor original reduced-order matrix {tilde over (G)}_D and the original-node impedance matrix X, and calculating load shifting distribution factor reduced-order matrix {tilde over (G)}_D,b1′

considering the disconnected branch b₁ according to the correction matrix {tilde over (G)}_D′ and

the load shifting distribution factor original reduced-order matrix {tilde over (G)}_D;

2.5) calculating generator shifting distribution factor reduced-order matrix {tilde over (G)}_G,b1′

considering the disconnected branch b₁ according to the load shifting distribution factor reduced-order matrix {tilde over (G)}_D,b1′

;

2.6) obtaining the load shifting distribution factor matrix G_D,b1 according to the load shifting distribution factor reduced-order matrix {tilde over (G)}_D,b1′ and

the relaxation node î

, and obtaining the generator shifting distribution factor matrix G_G,b1 according to the generator shifting distribution factor reduced-order matrix {tilde over (G)}_G,b1′ and

the relaxation node î

;

2.7) regarding the node impedance matrix X_b1 considering the disconnected branch b₁, the generator shifting distribution factor reduced-order matrix {tilde over (G)}_G,b1′

, and the load shifting distribution factor reduced-order matrix {tilde over (G)}_D,b1′

as new original-node impedance matrix X, new load shifting distribution factor original reduced-order matrix {tilde over (G)}_D, and new generator shifting distribution factor original reduced-order matrix {tilde over (G)}_G, and calculating the node impedance matrix, the load shifting distribution factor reduced-order matrix, and the generator shifting distribution factor reduced-order matrix and continuing the steps 2.2) to 2.7) until the disconnect branches in the d-th day are processed, and obtaining the generator shifting distribution factor matrix G_G,bB and the load shifting distribution factor matrix G_D,bB for the d-th day, and generating security constraint of the unit commitment model for the d-th day.

9. The non-transitory computer-readable storage medium of claim 6, wherein the step 3) comprises the following steps:

3.1) defining security constraint set Θ

of the power system as being equal to Ø

, Ø

is null set, and defining calculation loop variable s as being equal to zero, and calculating the no-security-constraint unit commitment model F≤

G_GP^T−

G_DL^T≤

F to obtain output X⁽⁰⁾ of each generator unit, and calculating the branch power flow F⁽⁰⁾ according to output X⁽⁰⁾ of each generator unit and the vector of the node loads L;

determining overload of each branch according to the branch power flow F⁽⁰⁾, upper limits of the branch power F, and lower limits of the branch flow F;

recording set of overloaded branches as Θ

⁽¹⁾;

if Θ

⁽¹⁾ is null set, going to step 3.3), and if Θ

⁽¹⁾ is not null set, updating the security constraint set Θ

=Θ

⁽¹⁾;

calculating security-constrained unit commitment model for the d-th day according the load shifting distribution factor matrix G_D,bB, the generator shifting distribution factor matrix G_G,bB, the upper limits of the branch power F, and the lower limits of the branch flow F;

solving the unit commitment model considering Θ

to obtain optimal solution p⁽¹⁾ of the unit commitment model considering Θ

in the first iteration;

3.2) defining s=s+1 to enter the next iteration;

for the s-th iteration, calculating the branch power flow F^(s) according to optimal solution P^(s) of the unit commitment model considering Θ

in the s-th iteration and the vector of the node loads L;

determining the overload of each branch according to the branch power flow F^(s), the upper limits of the branch power F, and the lower limits of the branch flow F, and recording the set of overloaded branches as Θ

^(s);

if Θ

^(s) is null set, going to the step 3.3), and if Θ

^(s) is not null set, updating the security constraint set Θ

=Θ

<sup>(s−

1)</sup>;

calculating the security-constrained unit commitment model for the d-th day;

solving the unit commitment model considering Θ

to obtain optimal solution P^(s+1) of the unit commitment model considering Θ

in the (s+1)-th iteration, and repeating the step 3.2);

3.3) considering the optimal solution P^(s) of the unit commitment model considering Θ

in the s-th iteration as the optimal solution of the security-constrained unit commitment model, and defining d=d+1 to go to the step 2.1) to calculate the operation simulation for the (d+1)-th day.