System parameter estimation method and estimation device

Abstract

To reduce an influence exerted on a power system for estimating a system parameter.SOLUTION: A system parameter estimation method is the method for estimating an apparent system parameter when a power system is viewed from an inverter linked with the power system at a linkage point. When P is considered as active power output from the inverter, Q is considered as reactive power output from the inverter, V is considered as voltage at the linkage point, Pa is considered as a target value of P, and Qa is considered as a target value of Q determined according to the target value of Pa, the method varies the active power P to the target value Pa and varies the reactive power Q to the target value Qa on providing a time difference between variation start timing of the active power P and variation start timing of the reactive power Q, and estimates the system parameter by a linear least-square method using (P, Q, V) in variation of the active power P or the reactive power Q.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a system parameter estimation method and an estimation device. [Background technology]

[0002] Conventionally, there has been known a method of estimating the impedance of an existing power system by varying the active power and reactive power output to a point of interconnection with the existing power system that supplies power to a load (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 4371062 Summary of the Invention [Problem to be solved by the invention]

[0004] To estimate system parameters such as impedance of a power system, it is necessary to vary both active and reactive power. However, if active and reactive power are varied to estimate system parameters, there is a risk that this will have a large impact on the power system, such as causing large disturbances.

[0005] The present disclosure provides a system parameter estimation method and estimation device that can reduce the impact on the power system due to the estimation of system parameters. [Means for solving the problem]

[0006] The present disclosure provides: A method for estimating apparent system parameters when viewing a power system from an inverter connected to the power system at a connection point, comprising: When P is the active power output from the inverter, Q is the reactive power output from the inverter, V is the voltage at the interconnection point, Pa is the target value of P, and Qa is the target value of Q determined according to the target value of Pa, A time lag is provided between the start timing of the change in the active power P and the start timing of the change in the reactive power Q, and then the active power P is changed to a target value Pa and the reactive power Q is changed to a target value Qa. A system parameter estimation method is provided, which estimates the system parameters by a linear least squares method using (P, Q, V) while changing the active power P or the reactive power Q.

[0007] The present disclosure also provides: An estimation device that estimates apparent system parameters when viewing a power system from an inverter connected to the power system at a connection point, the estimation device comprising: When P is the active power output from the inverter, Q is the reactive power output from the inverter, V is the voltage at the interconnection point, Pa is the target value of P, and Qa is the target value of Q determined according to the target value of Pa, A time lag is provided between the start timing of the change in the active power P and the start timing of the change in the reactive power Q, and then the active power P is changed to a target value Pa and the reactive power Q is changed to a target value Qa. The present invention provides an estimating device having an estimating circuit that estimates the system parameters by a linear least squares method using (P, Q, V) while the active power P or the reactive power Q is changing. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to reduce the impact on the power grid due to the estimation of grid parameters. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 2 is a diagram showing a state in which a grid-connection inverter is connected to a power grid. [Figure 2] FIG. 10 is an equivalent diagram for explaining a method for estimating system parameters. [Figure 3] FIG. 10 is a diagram for explaining an example of a method for changing active power and reactive power. [Figure 4] FIG. 10 is a diagram for explaining an example of a method for changing active power and reactive power. [Figure 5]FIG. 10 is a diagram for explaining an example of a method for changing active power and reactive power. [Figure 6] FIG. 10 is a diagram for explaining an example of a method for changing active power and reactive power. [Figure 7] FIG. 10 is a diagram for explaining an example of a method for changing active power and reactive power. [Figure 8] FIG. 10 is a diagram illustrating an example of a method for correcting R and X. [Figure 9] FIG. 10 is a diagram showing output waveforms during simulation of active power and reactive power. [Figure 10] FIG. 10 is a diagram illustrating an example of an estimation result of a grid voltage. [Figure 11] FIG. 10 is a diagram showing an example of the estimation results of R and X. [Figure 12] FIG. 10 is a diagram showing an example of the estimation results of R and X (one-time correction). [Figure 13] FIG. 10 is a diagram showing an example of the estimation results of R and X (corrected twice). DETAILED DESCRIPTION OF THE INVENTION

[0010] The following describes the embodiments.

[0011] <System configuration> 1 is a diagram showing a state in which a grid-connection inverter 5 is connected to a power system 1. When a grid-connection inverter 5 is connected to a power system 1, the actual power system 1 includes a large number of generators or loads.

[0012] The estimation device 7 according to this embodiment estimates apparent system parameters, which are how the power system 1 looks when viewed from the inverter 5 in a state where the inverter 5 is interconnected to the power system 1 at the interconnection point 3. Specific examples of apparent system parameters include the system impedance Z (=R+jX) of the power system 1, the voltage of the power system 1 (system voltage V gWhen the system impedance Z is expressed as a complex number such as R+jX, the real part is called the electrical resistance R of the power system 1, and the imaginary part is called the reactance X of the power system 1. The estimation device 7 is a device that can estimate these apparent system parameters using a simple method and suppress the estimation error.

[0013] The estimation device 7 has an estimation circuit 7a that estimates apparent system parameters when the power system 1 is viewed from an inverter 5 that is connected to the power system 1 at a connection point 3. The estimation circuit 7a is an electronic circuit that includes, for example, a CPU (Central Processing Unit), an FPGA (Field Programmable Gate Array), or an ASIC (Application Specific Integrated Circuit).

[0014] The estimation device 7 or estimation circuit 7a performs the various control operations described in this specification by a processor implemented as an electronic circuit executing a program stored in a memory implemented as an electronic circuit, or by being a circuit designed for a specific purpose.

[0015] <Method for estimating system parameters by fitting> Fig. 2 is an equivalent diagram for explaining a method for estimating system parameters. P is the active power output from the inverter 5 to the interconnection point 3, Q is the reactive power output from the inverter 5 to the interconnection point 3, V is the voltage at the interconnection point 3 (interconnection point voltage), and δ is the system voltage V g is the phase difference of the interconnection point voltage V relative to

[0016] In this case, complex power is expressed as the product of the phasor-expressed voltage and the complex conjugate of the phasor-expressed current, so

[0017]

number

[0018] If we separate Equation 1 into real and imaginary parts, we get

[0019]

number

[0020] Since the phase difference δ when looking at the power grid 1 from the inverter 5 is unknown, sin 2 δ+cos 2 Eliminating the phase difference δ using the equation δ=1, we get

[0021]

number

[0022] Rearranging equation 4, we get

[0023]

number

[0024] W=V 2 ,W g =V g 2 Then, Equation 5 becomes:

[0025]

number

[0026] On the left side of this equation 5A, if the third term is sufficiently small compared to the first and second terms, then approximately:

[0027]

number

[0028] The estimation device 7 according to this embodiment calculates R, X, and W. g This approximation formula is used to estimate the system parameters (i.e., R, X, and V) by fitting W as a linear equation of P and Q (α×P×β×Q+γ). g =√W g ) is estimated. α, β, γ are coefficients (α=2R, β=2X, γ=W g =V g 2 ).

[0029] When fitting is performed in the form of Equation 6, it is required to prepare three or more different data points (P, Q, V) (V = √W). However, even if there are three or more points, if (P, Q) are in a linear relationship (the power factor is fixed), three or more independent equations cannot be obtained, and fitting is not possible. For this reason, three or more different data points that do not result in a constant power factor are used for fitting. In other words, this condition is met when the rank of the matrix in which the vectors (P, Q, 1) are arranged is three.

[0030] The power factor is P / (√(P 2 +Q 2 ))".

[0031] <How to change the active power P and reactive power Q> As mentioned above, estimating system parameters requires that inverter 5 output active power P and reactive power Q at multiple power factors. It is easy to imagine that intentionally fluctuating the output power in order to estimate system parameters would be inconvenient for system operators. On the other hand, inverter 5, which is connected to power system 1 and outputs active power P and reactive power Q, normally controls the power factor to be constant. If constant power factor control is performed strictly, including during transient states in which active power P or reactive power Q is changing, system parameters cannot be estimated even using multiple pieces of data (P, Q, V) obtained while active power P or reactive power Q is changing.

[0032] Therefore, as shown in Fig. 3, the estimation device 7 according to this embodiment sets a non-zero time difference ΔT between the timing at which the active power P starts to change and the timing at which the reactive power Q starts to change, and then changes the active power P to a target value Pa and the reactive power Q to a target value Qa. At this time, on the premise that the inverter 5 performs constant power factor control in which the power factor is not 1, the estimation device 7 changes the target value Pa of the active power P or the target value Qa of the reactive power Q based on the constant power factor control. The estimation device 7 determines the target value Qa according to the target value Pa so that the power factor of the power output from the inverter 5 is constant.

[0033] By varying the active power P and reactive power Q with such a time difference ΔT, the power factor changes temporarily while the output of the active power P or reactive power Q is changing, providing an opportunity to collect multiple pieces of data (P, Q, V) for estimating system parameters. In the period after the active power P reaches the target value Pa and the reactive power Q reaches the target value Qa, the power factor returns to the constant value it had before the changes in the active power P and reactive power Q. Therefore, since the period during which the power factor changes to estimate the system parameters is temporary, the impact of large disturbances, etc., on the power system 1 can be mitigated.

[0034] For example, the estimation device 7 changes the target values ​​Pa and Qa, and then changes the active power P and reactive power Q in a ramp-like manner. This suppresses excessive changes in the active power P and reactive power Q or the power factor, thereby reducing the impact of large disturbances and the like on the power system 1. Note that the estimation device 7 may change the active power P and reactive power Q in a manner other than a ramp-like manner (for example, in a step-like manner).

[0035] 3 shows an example in which the change in the active power P starts first, but the change in the reactive power Q may start first. Which power should be changed first and how much time difference ΔT should be provided will be explained later. <System parameter estimation method using linear least squares method> Next, a specific example of fitting will be described.

[0036] The data of the i-th point (active power P, reactive power Q, interconnection point voltage V) used in the estimation by the above fitting is (P i ,Q i ,V i ) (where V i =√W i ). Also, y i =V 2 =W.

[0037] Error (y i -(α×P i +β×Q i +γ)) square sum E is

[0038]

number

[0039] To find α, β, and γ that minimize E,

[0040]

number

[0041] Equation 8 can be expressed as a determinant:

[0042]

number

[0043] The estimation device 7 calculates three or more different data points (P i ,Q i ,V i ) to calculate Equation 9 (where V i =√W i ), 2R, 2X and W gTherefore, the estimation device 7 estimates the α, β or γ (2R, 2X or W g ) to calculate the system parameters (R, X or V g ) can be easily estimated.

[0044] The estimation device 7 may, for example, estimate α, β, or γ (or R, X, or V g ) and adjusts the control conditions or output of the inverter 5. The estimation device 7 uses the estimated α, β, or γ (or R, X, or V g ) may be used to derive the power factor of the inverter 5.

[0045] <Calculation method of formula 9> In Equation 9, the Σ values ​​of the matrix elements and vector elements are calculated using the time series data (P i ,Q i ,W i (=V i 2 )) is the sum of the results of the operations on

[0046] The estimator 7 resets each Σ value of the matrix elements and vector elements in Equation 9 before starting the output change of the active power P and reactive power Q, and then starts the output change of the active power P or reactive power Q. During the output change of the active power P or reactive power Q, the estimator 7 adds the currently sampled (P i ,Q i ,W i (=V i 2 )) to derive the current value of each Σ value. i ,Q i ,W i (=V i 2 )) can be sequentially stored, thereby reducing the memory capacity.

[0047] After the output changes of the active power P and the reactive power Q are completed, the estimation device 7 completes the calculation of each Σ value of the matrix elements and vector elements in Equation 9. Therefore, even if the calculation processing capacity of the estimation circuit 7a is limited, the estimation device 7 can easily calculate the system parameters by the linear least squares method using Equation 9.

[0048] <Examples of multiple methods for changing the active power P and reactive power Q> The estimation device 7 may determine the time difference ΔT ( FIG. 3 ) at which the changes in the active power P and the reactive power Q begin, depending on the situation between the inverter 5 and the power grid 1. The estimation device 7 may determine which of the active power P and the reactive power Q begins to change first, depending on the situation between the inverter 5 and the power grid 1. If the time difference ΔT is too small, the estimation accuracy of the system parameters may deteriorate. If the time difference ΔT is too large, the interconnection point voltage V may temporarily fluctuate significantly while the active power P or the reactive power Q is changing, depending on whether the active power P or the reactive power Q is changed first.

[0049] For example, the estimation device 7 starts changing either the active power P or the reactive power Q, whichever starts changing later, when the change that started earlier has completed, or when the interconnection point voltage V deviates from a predetermined allowable range. By starting the change of either the active power P or the reactive power Q, whichever starts changing later, after the point at which the interconnection point voltage V deviates from the predetermined allowable range, it is possible to achieve as large a time difference ΔT as possible while keeping the fluctuation range of the interconnection point voltage V within a limited range. If the fluctuation range of the interconnection point voltage V is within a limited range, starting the change of either the active power P or the reactive power Q, which starts changing later, when the change that started earlier has completed, would not be expected to improve estimation accuracy even if a longer time difference ΔT is set. Therefore, the estimation device 7 can determine an appropriate time difference ΔT while ensuring estimation accuracy.

[0050] When the voltage V before the active power P or the reactive power Q starts to change is equal to or greater than the nominal value of the voltage V, the estimation device 7 may change the reactive power Q first when increasing the active power P (see FIG. 4). In this case, since the voltage V drops during the time difference ΔT, decreasing the reactive power Q first makes it less likely that the voltage V will deviate from a predetermined allowable range during the time difference ΔT. The nominal value of the voltage V is, for example, the center value of the predetermined allowable range of the voltage V.

[0051] When the voltage V before the active power P or the reactive power Q starts to change is equal to or greater than the nominal value of the voltage V, the estimation device 7 may change the active power P first when reducing the active power P (see FIG. 5). In this case, the voltage V drops during the time difference ΔT, so reducing the active power P first makes it less likely that the voltage V will deviate from a predetermined allowable range during the time difference ΔT.

[0052] When the voltage V before the active power P or the reactive power Q starts to change is equal to or lower than the nominal value of the voltage V, the estimation device 7 may change the active power P first when increasing the active power P (see FIG. 6). In this case, the voltage V increases during the time difference ΔT, so increasing the active power P first makes it less likely that the voltage V will deviate from a predetermined allowable range during the time difference ΔT.

[0053] When the voltage V before the start of the change in the active power P or the reactive power Q is equal to or lower than the nominal value of the voltage V, the estimation device 7 may change the reactive power Q first when reducing the active power P (see FIG. 7). In this case, since the voltage V increases during the time difference ΔT, increasing the reactive power Q first makes it less likely that the voltage V will deviate from a predetermined allowable range during the time difference ΔT.

[0054] <Correction method to reduce estimation error> In the estimation method using the linear least squares method described above, the estimation error in R and X is smaller than the estimation error in the system voltage V g The estimation error is not very large.

[0055] Therefore, the estimation device 7 estimates the system voltage V g may be used as is, and R and X may be corrected by the Newton method according to the following equation 10 and FIG.

[0056]

number

[0057] In order to enable correction using Equation 10, it is preferable that the estimation device 7 calculates a plurality of Σ values ​​shown in Equation 10 while the active power P or reactive power Q is changing.

[0058] 8 is a diagram showing an example of a method for correcting R and X. Since the correction accuracy (estimation accuracy) may be insufficient with a single correction, the estimation device 7 may correct (estimate) R and X by iterative calculation in which the correction is repeated two or three times.

[0059] The reason why R and X approach their true values ​​by the correction using Equation 10 will be explained below.

[0060] W g It is assumed that the estimated values ​​are close to the true values. In this case, the true R, X, W g Regarding the above, the square value W of the interconnection point voltage V when the active power P or reactive power Q is output satisfies the above equation 5A (if there is no noise).

[0061] For the measured data, the following f is minimized (R, X, W g ) is the most likely estimate.

[0062]

number

[0063] In (R,X) where f is minimum, "∂f / ∂R=0" and "∂f / ∂X=0" are satisfied. On the other hand, for (R,X) before correction, "∂f / ∂R=0" and "∂f / ∂X=0" are not satisfied.

[0064] In the neighborhood of the uncorrected estimate, ∂f / ∂R and ∂f / ∂X are

[0065]

number

[0066] dR and dX are calculated so that the left side of the equation becomes zero, and correction is performed using the calculated dR and dX. Here, T=P 2 +Q 2 Let's say.

[0067]

number

[0068]

number

[0069]

number

[0070] <Simulation example> Next, an example of the results of simulating the above estimation method will be shown.

[0071] In this simulation, the PU values ​​are set to a reference voltage of 6.6 kV and a reference capacity of 10 MVA. g = 0.95 pu (power system 1). The true value of R is 0.1 pu, and the true value of X is 0.2 pu. The active power P is varied from 0 to 0.5 pu for 1 second from time 0.5, and the reactive power Q is varied from 0 to -0.25 pu for 1 second, starting 0.5 seconds after the active power P starts to change (Figure 9). In this case, during the period from the start of the change in active power P to the completion of the change in reactive power Q, the estimator 7 acquires (P, Q, V) and calculates each of the above Σ values ​​(at a calculation interval of 1 millisecond). After the change in reactive power Q is completed, the estimator 7 uses these Σ values ​​to output the initial estimates using the linear least squares method and the corrected R and X.

[0072] 10 to 13 are diagrams showing examples of simulation results of the above estimation method. In FIGS. 10 to 13, the estimated values ​​of R and X are 0 and V until the first estimation is completed. g The estimated value is shown as 1 pu.

[0073] Figure 10 shows the system voltage V estimated by the above estimation method without correction by Equation 10. g As shown in Fig. 10, the estimated results of the system voltage V g The estimated value is almost equal to the true value.

[0074] 11 shows the estimation results of R and X estimated by the above estimation method without correction using Equation 10. If the deviation of the estimated values ​​of R and X from their true values ​​is not acceptable, correction using Equation 10 may be performed.

[0075] Fig. 12 shows the estimation results (correction results) of R and X obtained by one correction using Equation 10 and Fig. 8. Fig. 13 shows the estimation results (correction results) of R and X obtained by two corrections using Equation 10 and Fig. 8. The more the number of corrections is increased, the smaller the deviation of the estimated values ​​of R and X from the true values ​​is.

[0076] Although the embodiments have been described above, they are presented as examples and the present invention is not limited to the above embodiments. The above embodiments can be implemented in various other forms, and various combinations, omissions, substitutions, modifications, etc. can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as set forth in the claims. [Explanation of symbols]

[0077] 1 Power system 3 interconnection points 5 inverters 7 Estimation device

Claims

1. A method for estimating apparent system parameters when viewing a power system from an inverter connected to the power system at a connection point, comprising: When P is the active power output from the inverter, Q is the reactive power output from the inverter, V is the voltage at the interconnection point, Pa is the target value of P, and Qa is the target value of Q determined according to the target value of Pa, A time lag is provided between the timing at which the active power P starts to change and the timing at which the reactive power Q starts to change, and then the active power P is changed to a target value Pa and the reactive power Q is changed to a target value Qa; A system parameter estimation method, in which the system parameters are estimated by a linear least squares method using (P, Q, V) while the active power P or the reactive power Q is changing.

2. 2. The system parameter estimation method according to claim 1, wherein the change of the active power P or the reactive power Q, whichever starts to change later, starts when the interconnection point voltage V deviates from a predetermined range.

3. 3. The system parameter estimation method according to claim 2, wherein the change of the active power P or the reactive power Q that starts to change later is started when the change of the power that started to change earlier is completed.

4. 4. The system parameter estimation method according to claim 1, wherein, when the active power P is to be increased in a case where the voltage V before the active power P or the reactive power Q starts to change is equal to or higher than a nominal value of the voltage V, the reactive power Q is changed first.

5. 4. The system parameter estimation method according to claim 1, wherein when the active power P is to be reduced in a case where the voltage V before the active power P or the reactive power Q starts to change is equal to or higher than a nominal value of the voltage V, the active power P is changed first.

6. 4. The system parameter estimation method according to claim 1, wherein when the voltage V before the start of a change in the active power P or the reactive power Q is equal to or lower than a nominal value of the voltage V, the active power P is changed first when the active power P is to be increased.

7. 4. The system parameter estimation method according to claim 1, wherein, when the active power P is to be reduced in a case where the voltage V before the start of a change in the active power P or the reactive power Q is equal to or lower than a nominal value of the voltage V, the reactive power Q is changed first.

8. 4. The system parameter estimation method according to claim 1, wherein, during a period after the active power P reaches the target value Pa and the reactive power Q reaches the target value Qa, the active power P and the reactive power Q are changed so that the power factor returns to a constant value that was present before the changes in the active power P and the reactive power Q.

9. 4. The system parameter estimation method according to claim 1, wherein the active power P is changed in a ramp-like manner after a timing at which the active power P starts to change, and the reactive power Q is changed in a ramp-like manner after a timing at which the reactive power Q starts to change.

10. 4. The system parameter estimation method according to claim 1, wherein, while the active power P or the reactive power Q is changing, current values ​​of the matrix elements and the vector elements required for the calculation using the linear least squares method are added to previous values ​​of the matrix elements and the vector elements required for the calculation using the linear least squares method, and current sampling data required for the estimation using the linear least squares method is added to the previous values ​​of the matrix elements and the vector elements required for the calculation using the linear least squares method.

11. The system parameter estimation method according to claim 1 , wherein the system parameters are corrected by Newton's method.

12. An estimation device that estimates apparent system parameters when viewing a power system from an inverter that is connected to the power system at a connection point, comprising: When P is the active power output from the inverter, Q is the reactive power output from the inverter, V is the voltage at the interconnection point, Pa is the target value of P, and Qa is the target value of Q determined according to the target value of Pa, A time lag is provided between the timing at which the active power P starts to change and the timing at which the reactive power Q starts to change, and then the active power P is changed to a target value Pa and the reactive power Q is changed to a target value Qa; The estimating device has an estimating circuit that estimates the system parameters by a linear least squares method using (P, Q, V) while the active power P or the reactive power Q is changing.

Images