A bad data identification and state estimation method and device

Abstract

The present invention discloses a bad data identification and state estimation method and device. By constructing a measurement equation and using it to identify lever measurements and suspicious bad data measurements in power system data, and establishing a suspicious measurement set, and then identifying the suspicious measurement set by constructing a bad data identification model, the method and device can efficiently and accurately identify bad data in a multi-lever measurement system, and construct a power system state estimation model based on the bad data identification model. The power system state estimation model is used to perform performance estimation on an updated power data set that has eliminated bad data, thereby outputting an accurate power system state estimation and improving the reliability of the power system state estimation.

Description

Technical Field

[0001] The present invention relates to the technical field of power system state estimation, and in particular to a bad data identification and state estimation method and device. Background Art

[0002] State estimation is fundamental to achieving power system security assessment, real-time dispatch, and stable operation. The proportion of renewable energy connected to the power grid is increasing. Furthermore, the establishment of flexible DC systems has increased the degree of power electronics in the power grid. However, components such as on-load tap-changing transformers introduce a large number of non-smooth constraints, including discrete variables, limiters, and dead zones. These non-smooth constraints can affect the state estimation solution. Furthermore, the presence of bad data in the power grid can affect the accuracy of state estimation, reducing its convergence performance and even causing it to fail.

[0003] Currently, mixed-integer nonlinear programming methods are commonly used to optimize discrete variables and non-smooth constraints such as clipping and dead zones. However, this method is difficult to solve and has low computational efficiency. Furthermore, existing bad data detection and identification methods are mostly based on measurement residuals. If gross measurement errors occur in multi-lever measurement systems, they are difficult to detect based on the size of the residuals. This results in poor reliability for current power system state estimation. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a bad data identification and state estimation method and device, which can quickly and accurately identify bad data in a multi-lever measurement system and improve the reliability of power system state estimation.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A bad data identification and state estimation method comprises the steps of:

[0007] Constructing a measurement equation to identify leveraged measurements and suspicious bad data measurements in power system data through the measurement equation;

[0008] establishing a suspicious measurement set based on the leverage measurement and the suspicious bad data measurement;

[0009] Constructing a bad data identification model, and identifying the suspicious measurement set using the bad data identification model to obtain bad data;

[0010] Eliminating the bad data from the power system data to obtain an updated power data set;

[0011] A power system state estimation model is constructed based on the bad data identification model, and performance estimation is performed on the updated power data set using the power system state estimation model to obtain a power state estimation value.

[0012] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0013] A bad data identification and state estimation device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step in the bad data identification and state estimation method described above is implemented.

[0014] The beneficial effects of the present invention are: by constructing a measurement equation for identifying leverage measurements and suspicious bad data measurements in power system data, and establishing a suspicious measurement set, and then identifying the suspicious measurement set by constructing a bad data identification model, it is possible to efficiently and accurately identify bad data in a multi-lever measurement system, and construct a power system state estimation model based on the bad data identification model. The power system state estimation model is used to perform performance estimation on the updated power data set that eliminates the bad data, thereby outputting an accurate power system state estimation and improving the reliability of the power system state estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Flowchart of a bad data identification and state estimation method according to an embodiment of the present invention;

[0016] Figure 2 Schematic diagram of piecewise function smoothing of a bad data identification and state estimation method in an embodiment of the present invention;

[0017] Figure 3 Schematic diagram of the structure of a bad data identification and state estimation device in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] To illustrate the technical content, achieved objectives and effects of the present invention in detail, the following description is given in conjunction with the embodiments and accompanying drawings.

[0019] Please refer to Figure 1 , a bad data identification and state estimation method, comprising the steps of:

[0020] Constructing a measurement equation to identify leveraged measurements and suspicious bad data measurements in power system data through the measurement equation;

[0021] establishing a suspicious measurement set based on the leverage measurement and the suspicious bad data measurement;

[0022] Constructing a bad data identification model, and identifying the suspicious measurement set using the bad data identification model to obtain bad data;

[0023] Eliminating the bad data from the power system data to obtain an updated power data set;

[0024] A power system state estimation model is constructed based on the bad data identification model, and performance estimation is performed on the updated power data set using the power system state estimation model to obtain a power state estimation value.

[0025] From the above description, it can be seen that the beneficial effects of the present invention are: by constructing a measurement equation and using it to identify leverage measurements and suspicious bad data measurements in power system data, and establishing a suspicious measurement set, and then identifying the suspicious measurement set by constructing a bad data identification model, it is possible to efficiently and accurately identify bad data in a multi-lever measurement system, and construct a power system state estimation model based on the bad data identification model. The power system state estimation model is used to perform performance estimation on the updated power data set that eliminates bad data, thereby outputting an accurate power system state estimate and improving the reliability of the power system state estimate.

[0026] Furthermore, the construction of the measurement equation includes:

[0027] Obtain the topology and operating characteristics of the power system;

[0028] Obtaining a linearized expression of node injection power and a linearized expression of branch power according to the topological structure and operating characteristics;

[0029] The measurement equation is constructed according to the linear expression of the node injection power and the linear expression of the branch power.

[0030] From the above description, it can be seen that the linearized expression of node injection power and branch power is obtained through the topological structure and operating characteristics of the power system, and linear measurement equations are constructed according to the corresponding expressions, so that accurate measurement equations can be constructed to effectively identify leverage measurements and suspicious bad data measurements in power system data.

[0031] Furthermore, the identifying leveraged measurements and suspicious bad data measurements in the power system data by using the measurement equation includes:

[0032] Obtaining a Jacobian matrix according to the measurement equation;

[0033] identifying the leverage measure in the power system data using projection statistics and the Jacobian matrix;

[0034] The suspicious bad data measurement in the power system data is identified by using a maximum normalized difference method and the Jacobian matrix.

[0035] From the above description, it can be seen that the Jacobian matrix is ​​obtained through the measurement equation, and the Jacobian matrix is ​​combined with the projection statistics method and the maximum normalized difference method to identify leveraged measurements and suspicious bad data measurements in power system data, respectively. This can filter out the influence of measurement type identification of leveraged measurements and effectively screen out leveraged measurements and suspicious bad data measurements.

[0036] Furthermore, identifying the leverage measure in the power system data using projection statistics and the Jacobian matrix includes:

[0037] D i =(PS i / b i );

[0038]

[0039]

[0040] Where D i Used to indicate whether the corresponding measurement is a leverage measurement. If D i If it is greater than or equal to 1, the corresponding measurement is leverage measurement; PS i represents the projection value; b i represents the cutoff value; l i represents the transpose of the i-th row element in the Jacobian matrix H; v = l j (j=1,…,k), represents the j-th row element of the Jacobian matrix H; express and l j The degree of dispersion; τ is l i The dimension of .

[0041] From the above description, it can be seen that by adopting the projection statistics method and combining the Jacobian matrix to identify the leverage measurement in the power system, the leverage measurement in the power system can be accurately identified.

[0042] Furthermore, the identifying the suspicious bad data measurement in the power system data by using the maximum normalized difference method and the Jacobian matrix includes:

[0043]

[0044] Where r i N represents the normalized residual value of measurement i; R represents the diagonal matrix composed of measurement errors, is the variance of the error corresponding to measurement i; S represents the sensitivity matrix, S = IH (H T R -1 H)-1 H T R -1 , I is the identity matrix, H is the Jacobian matrix; r i Represents the residual value; for non-leveraged measurements, the normalized residual vector r N Obey the standard normal distribution, that is: r N ~N(0,1);

[0045] r N The maximum item in is compared with the statistical threshold ξ. If max(r N )>ξ, then the measurement value corresponding to the maximum residual term is the suspicious bad data measurement; the threshold ξ can be selected according to the required detection sensitivity level.

[0046] From the above description, it can be seen that by using the maximum normalized difference method and combining the Jacobian matrix to identify suspicious bad data measurements in the power system, suspicious bad data measurements in the power system can be accurately identified.

[0047] Furthermore, the constructing of the bad data identification model includes:

[0048] Establish the objective function of the bad data identification model;

[0049] Construct equilibrium node phase angle constraints, linearized zero injection equation constraints, and measurement equation constraints;

[0050] The objective function is constrained by the balance node phase angle constraint, the linearized zero injection equation constraint, and the measurement equation constraint to obtain the bad data identification model.

[0051] From the above description, it can be seen that by constructing the equilibrium node phase angle constraint, the linearized zero injection equation constraint and the measurement equation constraint to constrain the objective function of the bad data identification model, the recognition accuracy of bad data in the suspicious measurement set can be improved.

[0052] Furthermore, constraining the objective function by the balance node phase angle constraint, the linearized zero injection equation constraint, and the measurement equation constraint to obtain the bad data identification model includes:

[0053] The objective function of the bad data identification model is:

[0054]

[0055] The measurement equation constraints include:

[0056] stc L (x) = 0;

[0057]

[0058]

[0059] The equilibrium node phase angle constraint and the linearized zero injection equation constraint include:

[0060]

[0061] Where θ slack represents the equilibrium node phase angle; Represents the linearized measurement equation; M is a large positive number; σ i is the standard deviation of the measurement; d i is a 0 / 1 binary integer variable; when d i = 0, indicating the measurement value z in the suspicious measurement set i It is not bad data; on the contrary, when d i =1, indicating the corresponding measurement value z i is bad data; U, G, B, P and θ are parameters in the power system data.

[0062] From the above description, it can be seen that by restricting the objective function through specific measurement equation constraints, balance node phase angle constraints and linearized zero injection equation constraints, effective identification of bad data in the suspicious measurement set can be achieved.

[0063] Furthermore, constructing a power system state estimation model based on the bad data identification model includes:

[0064] Construct a mathematical model corresponding to the piecewise function of converter droop control in power systems;

[0065] Approximately processing the data model by a fitting function to obtain an approximate fitting function with smooth characteristics;

[0066] The power system state estimation model is obtained according to the bad data identification model and the approximate fitting function of the smooth characteristic.

[0067] From the above description, it can be seen that by processing the non-smooth characteristics existing in the converter model and using the fitting function to approximate the converter droop control piecewise function, the calculation efficiency and state estimation accuracy are greatly improved.

[0068] Furthermore, obtaining the power system state estimation model according to the bad data identification model and the approximate fitting function of the smooth characteristic includes:

[0069]

[0070] in, is the integer variable solution of the mixed integer linear programming bad data identification model, c(x) is the constraint condition in the bad data identification model; h i (x) measurement equation; z i represents the measurement value in the suspicious measurement set; R represents the diagonal matrix composed of measurement errors.

[0071] From the above description, it can be seen that the power system state estimation model is obtained based on the bad data identification model and the approximate fitting function of the smooth characteristics, so that the power system state estimation model can adapt to the state estimation performance of the AC / DC hybrid power system with non-smooth characteristics, thereby improving the reliability of the power system state estimation.

[0072] Please refer to Figure 3 Another embodiment of the present invention provides a bad data identification and state estimation device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the processor implements each step of the bad data identification and state estimation method described above.

[0073] The bad data identification and state estimation method and apparatus described above are suitable for bad data identification and state estimation with non-smooth characteristics. They can quickly and accurately identify bad data in a multi-lever measurement system and effectively handle the non-smooth constraints in the system, thereby facilitating rapid state estimation and improving the adaptability of state estimation to non-smooth characteristics. A specific embodiment is provided below to illustrate the method and apparatus.

[0074] Example 1

[0075] Please refer to Figure 1 A bad data identification and state estimation method, characterized by comprising the steps of:

[0076] S1. Constructing a measurement equation to identify leveraged measurements and suspicious bad data measurements in power system data through the measurement equation. The steps of constructing the measurement equation are as follows:

[0077] S11. Obtain the topology and operating characteristics of the power system. In order to obtain the linearization constraints driven by the physical model, the basic principle of linear approximation is obtained based on the actual topology and operating characteristics of the power grid:

[0078] 1) When the system is operating normally, the voltage amplitude of each node is close to the voltage reference value, and the per-unit value is expressed as: U i ≈1.0pu, i.e. 1.0pu of power frequency overvoltage;

[0079] 2) The phase difference between the first and last nodes of the line is small, and the approximate treatment is: cosθ ij ≈1, sinθ ij ≈θ ij; where θ ij =θ i -θ j , represents the phase angle difference between node i and node j;

[0080] S12. Obtain a linearized expression for node injection power and a linearized expression for branch power based on the topological structure and operating characteristics; apply the above principle to the nonlinear expression in the bad data identification model, as shown below:

[0081] Nonlinear measurement equation of node injection power:

[0082]

[0083] Nonlinear measurement equation of branch power:

[0084]

[0085]

[0086] Zero injection equality constrains nonlinear equations:

[0087]

[0088] Furthermore, the physical linearized expressions of the measurement equation and the zero injection equality constraint are obtained:

[0089] Linearized expression of node injection power:

[0090]

[0091] Where U i ,U j Represents the voltage amplitude of nodes i and j respectively; θ i ,θ j Represents the voltage phase angle of node i, j respectively; P i , Q i They represent the active and reactive power injection quantities of node i respectively; G ij 、B ij Represent the real part and imaginary part of the element in the i-th row and j-th column of the node admittance matrix, respectively, representing the conductance and admittance, respectively. The node admittance matrix is ​​calculated from the line impedance parameters and the node connection relationship; in the above formula, ΔP i , ΔQ i , represents the fitting error based on data driving, obtained by partial least squares regression fitting;

[0092] Branch power linearization equation:

[0093]

[0094]

[0095] Where: P ij , Q ij They represent the active and reactive power quantities at the head end of the branch, respectively, and the direction is from node i to node j; P ji , Q ji They represent the active and reactive power quantities at the end of the branch, respectively, and the direction is from node j to node i; g ij 、b ij Respectively represent the conductance and susceptance of the branch where node i and node j are located; in the above formula, ΔP ij , ΔQ ij , ΔP ji , ΔQ ji It represents the fitting error based on data-driven fitting, obtained by partial least squares regression fitting;

[0096] Linearized equations with zero injection equality constraints:

[0097]

[0098] In the above formula, ΔP i , ΔQ i , represents the fitting error based on data driving, obtained by partial least squares regression fitting; the data driving principle based on PLS regression is as follows:

[0099] ΔZ l =C l ·Z l +D l ;

[0100] ΔZ l =[ΔP ij ΔQ ij ΔP ji ΔQ ji ΔP i ΔQ i ];

[0101] Z l =[P d Q d ] T ;

[0102] Where C l , D l is the fitting coefficient, obtained by PLS fitting; P d , Q d are the non-zero active and reactive power column vectors respectively;

[0103] S13, constructing the measurement equation according to the linear expression of the node injection power and the linear expression of the branch power; obtaining the measurement equation according to the linear expression of the node injection power and the linear expression of the branch power;

[0104] After the measurement equation is constructed, identifying leveraged measurements and suspicious bad data measurements in power system data using the measurement equation includes the following steps:

[0105] S14, obtaining a Jacobian matrix according to the measurement equation;

[0106] S15. Identify the leverage measurement in the power system data using projection statistics and the Jacobian matrix, specifically:

[0107] The projection statistic PS i The calculation formula is as follows:

[0108]

[0109] Where, l i represents the transpose of the i-th row element in the Jacobian matrix H; v = l j (j=1,…,k), represents the j-th row element of the Jacobian matrix H; express and l j the degree of dispersion;

[0110] described The calculation formula is:

[0111]

[0112] lomed = [(k+1) / 2];

[0113] In the formula, “[]” means rounding down;

[0114] The projection statistic PS i With the cutoff value b i Compare and determine whether it is a leverage measurement; under Gaussian distribution, the projection statistic obeys the chi-square distribution with degrees of freedom τ, where τ is l i The dimension of ; where the cutoff value b is defined i as follows:

[0115]

[0116] represents a chi-square distribution with τ degrees of freedom and a probability of 0.975;

[0117] Compare the projection values ​​PS i With cutoff value bi :

[0118] D i =(PS i / b i )≥1;

[0119] Where D i Used to indicate whether the corresponding measurement is a leverage measurement. If D i If it is greater than or equal to 1, the corresponding measurement is a leverage measurement;

[0120] S16. Identifying the suspicious bad data measurement in the power system data using the maximum normalized difference method and the Jacobian matrix;

[0121] Divide the residual value by the corresponding diagonal term in the residual covariance matrix to obtain the residual vector r N The expression:

[0122]

[0123] Where r i N represents the normalized residual value of measurement i; R represents the diagonal matrix composed of measurement errors, is the variance of the error corresponding to measurement i; S represents the sensitivity matrix, S = IH (H T R -1 H) -1 H T R -1 , I is the identity matrix, H is the Jacobian matrix; r i Represents the residual value; for non-leveraged measurements, the normalized residual vector r N Obey the standard normal distribution, that is: r N ~N(0,1);

[0124] r N The maximum item in is compared with the statistical threshold ξ. If max(r N )>ξ, then the measurement value corresponding to the maximum residual term is the suspicious bad data measurement; the threshold ξ can be selected according to the required detection sensitivity level;

[0125] S2. Establishing a suspicious measurement set based on the leverage measurement and the suspicious bad data measurement; that is, establishing the corresponding suspicious measurement set based on the leverage measurement and the suspicious bad data measurement obtained in step S1;

[0126] S3. Build a bad data identification model, and identify the suspicious measurement set using the bad data identification model to obtain bad data;

[0127] S4, removing the bad data from the power system data to obtain an updated power data set;

[0128] S5. Constructing a power system state estimation model based on the bad data identification model, and performing performance estimation on the updated power data set using the power system state estimation model to obtain a power state estimation value.

[0129] Example 2

[0130] This embodiment differs from the first embodiment in that the method for constructing the bad data identification model is specifically defined;

[0131] The construction of the bad data identification model in step S3 includes the following steps:

[0132] S31. Establishing an objective function of a bad data identification model; the objective function of the bad data identification model is:

[0133]

[0134] S32, constructing equilibrium node phase angle constraints, linearized zero injection equation constraints, and measurement equation constraints;

[0135] The measurement equation constraints include:

[0136] stc L (x) = 0;

[0137]

[0138]

[0139] The equilibrium node phase angle constraint and the linearized zero injection equation constraint include:

[0140]

[0141] θ slack represents the equilibrium node phase angle; Represents the linearized measurement expression, namely the linearized expression of the node injection power and the branch power linearized expression; M is a large positive number; σ i is the standard deviation of the measurement; d i is a 0 / 1 binary integer variable; when d i = 0, indicating the measurement value z in the suspicious measurement set i It is not bad data; on the contrary, when d i =1, indicating the corresponding measurement value z i is bad data; U, G, B, P and θ are parameters in the power system data; they should be eliminated in the subsequent weighted least squares state estimation;

[0142] S33, constraining the objective function by the balance node phase angle constraint, the linearized zero injection equation constraint, and the measurement equation constraint to obtain the bad data identification model;

[0143] Please refer to Figure 2 Step S5 further includes: based on the bad data identification model, taking into account the non-smooth characteristics of the converter, constructing a weighted least squares state estimation model, and processing the non-smooth characteristics existing in the converter model, specifically:

[0144] S51. Construct a mathematical model corresponding to a piecewise function of converter droop control in a power system;

[0145] Mathematical modeling of the droop control piecewise function:

[0146]

[0147] Where, P max 、P min Represents the DC side active power P dc Upper and lower limits of P db Indicates that when the voltage U dc Active power in the dead zone; U dbh 、U dbl Respectively represent the upper and lower limits of the voltage dead zone; U h 、U l They correspond to the upper and lower bounds of the droop control linear function curve respectively; U max 、U min They represent the upper and lower bounds of the DC bus voltage amplitude respectively; α1 and α2 represent the droop control coefficients respectively, and α1 < 0, α2 < 0. The relationships between the droop control coefficients and the active power and voltage are:

[0148]

[0149] Among them, the control methods of the converter mainly include master-slave control, voltage margin control and droop control. In droop control, the VSC in the flexible DC system can automatically adjust the DC side power and voltage. Compared with master-slave control and voltage margin control, droop control can instantly adjust the power flow of each converter station without the need for real-time communication between converter stations, and has strong flexibility. The droop control characteristics taking into account the limit and dead zone can be characterized by piecewise functions. Due to the strong non-smooth characteristics of piecewise functions, when the system runs near the inflection point, the derivative is discontinuous, which easily leads to algorithm calculation failure. Therefore, it is necessary to solve the problems caused by the non-smoothness of piecewise functions.

[0150] S52, performing approximate processing on the data model by using a fitting function to obtain an approximate fitting function with smooth characteristics;

[0151] Use the fitting function to approximate the above piecewise function:

[0152]

[0153] Wherein, k is the fitting coefficient, and in an optional embodiment, k=500;

[0154] S53: Obtain the power system state estimation model according to the bad data identification model and the approximate fitting function of the smooth characteristic; the power system state estimation model is a weighted least squares state estimation model:

[0155]

[0156] stc(x)=0;

[0157] in, is the integer variable solution of the mixed integer linear programming bad data identification model, c(x) is the constraint condition in the bad data identification model; h i (x) measurement equation; z i Represents the measurement value in the suspicious measurement set; when measurement z i When it is bad data, the corresponding If it is 1, the weight of the corresponding residual term in the objective function is set to 0, ensuring that this term has no effect during the optimization process, thereby eliminating bad data.

[0158] Example 3

[0159] Please refer to Figure 3 A bad data identification and state estimation device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of a bad data identification and state estimation method as described in Embodiment 1 or 2 is implemented.

[0160] In summary, the present invention provides a bad data identification and state estimation method and device, which utilizes projection statistics and maximum normalized residual method to compress the suspicious measurement set, and linearizes the nonlinear measurement equations and equality constraints in the state estimation through data physical fusion driven linearization method, and constructs a mixed integer linearization bad data identification method for the suspicious measurement set; the droop control piecewise function contained in the converter is smoothed by the approximate fitting method, and on the basis of bad data identification, the minimum sum of squared residuals is used as the objective function to construct a weighted least squares state estimation model taking into account non-smooth characteristics; thereby, the influence of the bad data identification link and non-smooth characteristics on the state estimation in the power system can be taken into account, which is conducive to fast and effective bad data identification and state estimation, and is suitable for application in new power system state estimation applications with a high proportion of renewable energy.

[0161] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's description and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

[0162] It will be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention may be implemented in various computer languages, for example, the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0163] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0164] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0165] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0166] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0167] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A bad data identification and state estimation method, characterized in that: Including steps: Constructing a measurement equation to identify leveraged measurements and suspicious bad data measurements in power system data through the measurement equation; establishing a suspicious measurement set based on the leverage measurement and the suspicious bad data measurement; Constructing a bad data identification model, and identifying the suspicious measurement set using the bad data identification model to obtain bad data; Eliminating the bad data from the power system data to obtain an updated power data set; constructing a power system state estimation model based on the bad data identification model, and performing performance estimation on the updated power data set using the power system state estimation model to obtain a power state estimation value; The construction of the bad data identification model includes: Establish the objective function of the bad data identification model; Construct equilibrium node phase angle constraints, linearized zero injection equation constraints, and measurement equation constraints; The objective function is constrained by the balance node phase angle constraint, the linearized zero injection equation constraint, and the measurement equation constraint to obtain the bad data identification model, which includes: The objective function of the bad data identification model is: ； The measurement equation constraints include: ； ； ； The equilibrium node phase angle constraint and the linearized zero injection equation constraint include: ； represents the equilibrium node phase angle; represents the linearized measurement equation; is a positive number; is the standard deviation of the measurement; is a 0 / 1 binary integer variable; when When , it indicates the measurement value in the suspicious measurement set It is not bad data; on the contrary, when When , it indicates the corresponding measurement value It’s bad data; Representation node With node The phase angle difference; 、 Represents nodes respectively The active and reactive injected power to be determined; 、 They represent the node admittance matrix Row, No. The real and imaginary parts of the column elements represent conductance and susceptance respectively; Represents nodes respectively The voltage amplitude; The step of obtaining the power system state estimation model based on the bad data identification model and the approximate fitting function of the smooth characteristic includes: ； ； in, is the integer variable solution of the mixed integer linear programming bad data identification model, identifying constraints in a model for the bad data; measurement equations; Represents the measurement value in the suspicious measurement set; Represents the diagonal matrix composed of measurement errors.

2. The bad data identification and state estimation method according to claim 1, characterized in that: The construction of the measurement equation includes: Obtain the topology and operating characteristics of the power system; Obtaining a linearized expression of node injection power and a linearized expression of branch power according to the topological structure and operating characteristics; The measurement equation is constructed according to the linear expression of the node injection power and the linear expression of the branch power.

3. The bad data identification and state estimation method according to claim 1, characterized in that: The identifying of leveraged measurements and suspicious bad data measurements in the power system data by using the measurement equation includes: Obtaining a Jacobian matrix according to the measurement equation; identifying the leverage measure in the power system data using projection statistics and the Jacobian matrix; The suspicious bad data measurement in the power system data is identified by using a maximum normalized difference method and the Jacobian matrix.

4. The bad data identification and state estimation method according to claim 3, characterized in that: The identifying the leverage measure in the power system data using projection statistics and the Jacobian matrix includes: ； ； ； Where, Used to indicate whether the corresponding measurement is a leverage measurement. If it is greater than or equal to 1, the corresponding measurement is a leverage measurement; Represents the projection value; represents the cutoff value; represents the Jacobian matrix The Transpose of row elements; , represents the Jacobian matrix No. row elements; express and the degree of dispersion; yes Dimensionality; The degrees of freedom are , a chi-square distribution with a probability of 0.

975.

5. The bad data identification and state estimation method according to claim 3, characterized in that: The identifying the suspicious bad data measurement in the power system data by using the maximum normalized difference method and the Jacobian matrix includes: ； Where, Indicates measurement The normalized residual value of ; represents the diagonal matrix composed of measurement errors, It is measurement The variance of the corresponding error; represents the sensitivity matrix, , is the identity matrix, is the Jacobian matrix; Represents the residual value; for non-leveraged measurements, the normalized residual vector Obey the standard normal distribution, that is: ; Will The maximum item and statistical threshold in For comparison, if , then the measurement value corresponding to the maximum residual item is the measurement of the suspicious bad data; the threshold The choice can be made based on the desired level of detection sensitivity.

6. The bad data identification and state estimation method according to claim 1, characterized in that: The constructing of a power system state estimation model based on the bad data identification model includes: Construct a mathematical model corresponding to the piecewise function of converter droop control in power systems; Approximately processing the mathematical model by using a fitting function to obtain an approximate fitting function with smooth characteristics; The power system state estimation model is obtained according to the bad data identification model and the approximate fitting function of the smooth characteristic.

7. A bad data identification and state estimation device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, each step of the bad data identification and state estimation method according to any one of claims 1 to 6 is implemented.

Images