JP2645001B2 - Preventive control method of power system dynamic stability - Google Patents

Description

The present invention relates to a preventive control method for monitoring and controlling the dynamic stability of a power system online.

(Prior Art) As the power system becomes larger and more complex, it is increasingly important to grasp and maintain the stability of the power system. In such a situation, in order to ensure the stability of the power system in the future, not only the emergency stabilization control but also the normal preventive control will be important.

Until now, stability improvement measures applied to an actual system mainly consisted of an emergency control system for urgently controlling various stabilizing devices on the condition that a disturbance occurred. On the other hand, the so-called preventive control, which monitors the stability at normal times and performs necessary control in advance to maintain sufficient stability against a certain level of disturbance, is still in the research stage and is applied to actual systems. There are few examples.

By the way, in order to realize preventive control, it is essential to develop a higher-speed operation algorithm together with hardware technologies such as a high-speed data transmission system and a large-capacity and high-speed computer.

On the other hand, the power system is a large and complex system,
It is considered that the number of disturbances (prescribed disturbances) that can occur is an enormous number. Therefore, it is inefficient and impractical to determine the dynamic stability in detail by eigenvalue analysis or transient stability analysis for all the specified disturbances.

Also, in the process of calculating the preventive control amount, the eigenvalue calculation is performed by execution and error for all the assumed disturbances that become unstable.
Alternatively, it is not advisable to obtain an appropriate control target and control amount using a transient stability calculation program.

(Problems to be Solved by the Invention) In order to maintain the dynamic stability of the power system as described above, it is necessary to develop a high-speed operation algorithm. In this case, a more realistic and efficient preventive control method is used. In order to realize the above, (1) Study of a dynamic stability evaluation method that can quickly select an assumed disturbance that is likely to become unstable (2) It is necessary to study a method that can calculate an appropriate preventive control amount more quickly .

Therefore, the present invention uses an operation algorithm combining the eigenvalue sensitivity analysis method and the linear programming method that focuses on the change of the eigenvalue representing the system dynamics with respect to the change of the system parameter. An object of the present invention is to provide a preventive control method of power system dynamic stability that can be calculated.

[Structure of the Invention] (Means for Solving the Problems) In order to achieve the above object, the present invention evaluates the dynamic stability of the power system with respect to a preset assumed disturbance and performs necessary control in advance. In the preventive control method, dynamic stability against expected disturbance is evaluated by eigenvalue sensitivity analysis,
The generator output adjustment required to stabilize the eigenvalue determined to be unstable as a result is determined by the sensitivity coefficient of the eigenvalue real part with respect to the generator output and the maximum variation of the eigenvalue real part obtained from the dynamic stability evaluation results. This is characterized in that the dynamic stability of the power system is preventively controlled by obtaining a solution of the linear programming method.

(Operation) Therefore, in such a power system dynamic stability preventive control system, an efficient eigenvalue sensitivity analysis method is applied to the evaluation of dynamic stability against a large number of assumed disturbances, and it is determined that the system is unstable there. For the eigenvalues obtained, the sensitivity coefficient of the eigenvalue real part with respect to the generator output and the maximum change in the eigenvalue real part known from the results of the dynamic stability evaluation are introduced as preventive control variables necessary for stabilization. It can be obtained as a solution of simple linear programming.

Embodiment An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is an overall flow chart for explaining a preventive control system for power system dynamic stability according to the present invention.
First, in step 1, system operation state data indicating the current operation state of the system, that is, the generator output, voltage, etc., is collected. Then a number of assumptions disturbance f _1, f _2, which is prepared in advance in Step 2, performs efficient kinetic stability evaluation by eigenvalue sensitivity analysis in Step 3 with respect ...... f _n, the kinetics unstable assumed Select a disturbance. If it is determined that there is at least one unstable assumed disturbance due to the selection of the assumed disturbance in step 3, a linear programming is performed in step 4 so as to maintain the dynamic stability for all the assumed disturbances. The required preventive control amount is calculated by Systematic Tech. Then, it is determined whether or not it is necessary to calculate the transient stability with respect to the preventive control amount calculated in step 4. If it is determined that the calculation of the transient stability is necessary, the calculation of the transient stability is performed in step 6. Is executed to confirm the effect of performing the preventive control based on the preventive control amount. Here, however, it is sufficient to analyze the most unstable assumed disturbance.

Further, the transient stability calculated in step 6 and the preventive control amount determined to be unnecessary in step 5 are presented to the system operator in step 7 and reflected in the operation of the actual system. After a series of calculations is completed, after a lapse of a predetermined time or when the system operation state changes and the operator determines that it is necessary, the calculation of the present preventive control method is restarted.

The above is the overall flow of the preventive control method. Here, the dynamic stability evaluation part by the eigenvalue sensitivity analysis method in step 3 and the calculation part of the preventive control amount by the linear programming method in step 4 are features of the present invention. The specific contents will be described.

First, dynamic stability evaluation by eigenvalue sensitivity analysis will be described. As a system model for dynamic stability analysis, a linear model that linearly approximates small changes near the operating point is used. As a method of determining the stability of the model, an eigenvalue analysis method has been conventionally used. However, if the system dimension is large and the number of analysis cases is large, the calculation time is too long, so this method cannot be used directly for online calculation.

On the other hand, a sensitivity analysis method for eigenvalues has been proposed as a method for analyzing how eigenvalues of a target system change when system parameters change (for example, JEVan).
Nees et al. “Sensitivities of Large-Loop control System
s ", IEEE Trans on Automatic control, Vol. 10, 308-
315 pages, July, 1965).

According to this method, the eigenvalue calculation basically needs to be performed only once for one operating point. The change [Δλ] of the eigenvalue with respect to the change of the parameter is as follows: (a) the differential coefficient [∂A / ∂γ] regarding the parameter γ of the coefficient matrix [A] (b) the eigenvector [W] of [A] and the transposed matrix [A]
Since the calculation can be performed using the eigenvector [V] of ^t , the influence of the parameter can be obtained relatively easily.

In order to apply this method to dynamic stability evaluation for each assumed disturbance, processing as shown in FIG. 2 is performed. in this case,
As a specified disturbance for dynamic stability evaluation, the opening of a transmission line is the main one, and here the parameter is used as a system impedance.

In FIG. 2, steps 11 to 14 are a part of the preparation calculation of the sensitivity analysis. That is, in step 11, a power flow calculation for determining the current power flow distribution state is performed, and based on the result of the power flow calculation, in step 12, the system coefficient matrix [A] ₀
Create Then the eigenvalue .lambda.i ₀ respectively from eigenvalue calculation in Step _{13 [A] 0, i =} 1~n (n Dimension) eigenvector for each eigenvalue [Wi], and i = 1 to n, transpose matrix ^{[A] t} Eigenvector [Vi], i = 1 to n.

Steps 15 to 22 are the part of the sensitivity analysis that is repeatedly calculated for each assumed accident case. That is, step 15 is a part for selecting one of the predetermined assumed accident cases (open route) from the file. Step 16 is a part for performing a power flow calculation for determining a power flow distribution state after the route is opened, which is the expected disturbance. In step 17, a coefficient matrix [A] 'of the system after the line is opened is created based on the result of the power flow calculation in step 16. Next, at step 18, a change [ΔA] of the coefficient matrix [A] accompanying the route opening is calculated. Using this, in step 19, a predicted change value Δλi, i = 1 to n of the eigenvalue λi, i = 1 to n is calculated. When this Δλi is obtained, the expected value λi of the eigenvalue after the line is opened can be calculated in step 20, and the dynamic stability is evaluated in step 21 using this λi. That is, if the real part of the eigenvalue λi is positive, the vibration component is unstable, and if the real part is negative, it can be determined that it is stable. If all the assumed accident cases are not completed in step 22, return to step 15 and step 15
Repeat up to 21. Second if all assumed accident cases are completed
The dynamic stability evaluation of the figure is completed.

Next, calculation of the preventive control amount by the linear programming method will be described.

As a result of the stability determination against the assumed disturbance, the dynamic stability cannot be maintained under the current system operation conditions, that is, when it is determined that there is one or more eigenvalues whose real part is positive as a result of the calculation in step 3 of FIG. Then, it is to be stabilized by adjusting the output of the generator as preventive control.

The preventive control amount (generator output adjustment amount) when there are N eigenvalues of the system exhibiting dynamic stability is determined based on the following concept.

When the real part of the eigenvalue i of the system changes by Δσ due to the output adjustment ΔPj of the generator j, the following equation generally holds.

Here, i: eigenvalue number, j: adjusted generator number (m units) Then, as the preventive control of the dynamic stability, the eigenvalue real part σ is always calculated for all N eigenvalues determined to be unstable.
This results in finding (ΔPj, i = 1 to m) such that i changes from positive to negative. However,
At this time, the following conditions shall be applied.

(A) To maintain the power supply and demand balance of the entire system.

(B) Make the total amount of power adjustment as small as possible.

(C) Consider the adjustment width determined for each generator.

In order to facilitate mathematical processing in the linear programming, the output adjustment amount ΔPj is described as follows.

Here, ΔPj ⁺ : an increase in the output adjustment amount of the generator j ΔPj ⁻ : a decrease in the output adjustment amount of the generator j, ΔPj ⁺ , ΔPj ⁻ ≧ 0 is always obtained.

As a solution to such a problem, linear programming can be applied as it is. That is, Function under To find ΔPj ⁺ and ΔPj ⁻ that minimize.

Here, Δσi ^MAX , i = 1 to N in the equation (5) is the maximum change amount (positive direction) of the real part of the eigenvalue i in all the assumed accident cases, which is the step in FIG. 3 can be obtained in the process of dynamic stability evaluation by the eigenvalue sensitivity analysis method. Here, FIG. 4 shows an example of a dynamic stability evaluation result by eigenvalue sensitivity analysis.

The figure shows the change direction and stability of the real part of the eigenvalue of the power fluctuation mode for each assumed accident (transmission line opening).

Of the 23 assumed accidents, the dynamic branch became unstable in a total of 9 cases with open branch numbers 64, 63, 61, 60, 58, 44, 20, 38, and 39. These 9 cases have eigenvalues No. 73 and No. 85
Are unstable in the two oscillation modes (N = 2). The maximum positive variation in the real part of these two eigenvalues is +1.076 in the case of open branch No. 38 for No. 73, and the same for open branch No. 38 for No. 85. In the case of +0.0324 is given.

In the equation (6), ωj ⁺ and ωj ⁻ are weighting factors in the increasing and decreasing directions of the output of the generator j, respectively, and may be set in advance based on the adjustment order of the generator.

FIG. 3 shows the entire flow of the process of calculating the preventive control amount by the linear programming method described above. That is,
As shown in FIG. 3, in step 101, Δσi ^MAX is determined, and in step 102, the sensitivity coefficient αij is calculated. Step 103 is a part for calculating the output adjustment amount ΔPj of the generator by applying the linear programming based on the conditions formulated by the equations (3) to (6). Step 104 is performed when it is found that, as a result of the output adjustment amount ΔPj calculated in step 103, the eigenvalue that was initially in the stable region (the real part is negative) enters the unstable region (the real part is positive). Returning to step 103 again indicates that the output adjustment amount ΔP is recalculated. In this case, redefine the other was added with Derutashigumaai ^MAX for eigenvalues can become a new unstable (5) of Derutashigumaai ^MAX for initially unstable eigenvalue may be determined to ΔPj again.

As described above, in the present embodiment, if the steps 3 and 4 shown in FIG. 1 are configured as shown in FIGS. 2 and 3, respectively, the eigenvalue of the system with respect to a preset assumed disturbance (open line) is set. Since the change can be grasped, an efficient dynamic stability evaluation with a high calculation speed can be realized. Further, the preventive control amount (generator output adjustment amount) required to stabilize the eigenvalue determined to be unstable as a result of this evaluation can be uniquely determined by applying the linear programming.

[Effects of the Invention] As described above, according to the present invention, an efficient eigenvalue sensitivity analysis method is applied to the evaluation of dynamic stability against a large number of assumed disturbances, and stable eigenvalues determined to be unstable there are obtained. The preventive control amount (generator output adjustment amount) required to make the system is introduced by introducing the sensitivity coefficient of the eigenvalue real part to the generator output and the maximum change amount of the eigenvalue real part known from the results of dynamic stability evaluation. Since it is obtained as a solution of a systematic linear programming method, it is possible to realize a high-speed operation algorithm for maintaining dynamic stability, and it is possible to select expected disturbance at high speed and calculate an appropriate preventive control amount A preventive control method for power system dynamic stability can be provided.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is an overall flow chart showing an embodiment for explaining a preventive control method of power system dynamic stability according to the present invention, FIG.
FIG. 3 is a flowchart showing the specific contents of the dynamic stability evaluation part by the eigenvalue sensitivity analysis method in FIG. 1, and FIG. 3 shows the specific contents of the preventive control amount calculation part by the linear programming method in FIG. FIG. 4 is a flowchart for explaining the procedure of FIG. Step 1 ... System status data collection unit, Step 2 ...
Assumed disturbance data section, Step 3 ... Dynamic stability evaluation section by eigenvalue sensitivity analysis, Step 4 ... Preventive control amount calculation section by linear programming, Step 5 ... Determination section whether stability calculation is necessary , Step 6: Transient stability calculation unit,
Step 7: Control execution unit by system operator.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Hideki Fujita 1 at 20 Kitakanyama, Odaka-cho, Midori-ku, Nagoya-shi Within Chubu Electric Power Co., Inc. (72) Inventor Isamu Suzumura 1 Higashi-shinmachi, Higashi-ku, Nagoya (72) Inventor Tokuhiro Sugiura, 18 Yanagicho, Nagano City, Chubu Electric Power Co., Inc., Nagano Branch (72) Inventor, Kaoru Koyanagi 1 Toshiba Town, Fuchu City, Tokyo, Japan Higashishiba Fuchu Plant ( 72) Inventor: Shigeru Sato 1-1-1, Shibaura, Minato-ku, Tokyo, Japan Inside the Toshiba Corporation headquarters office (72) Inventor Toshio Shimoda 1-1-1, Shibaura, Minato-ku, Tokyo, Tokyo Toshiba Corporation office ( 56) References JP-A-60-223437 (JP, A) JP-A-54-120839 (JP, A)

Claims (1)

(57) [Claims] 1. A preventive control method for evaluating dynamic stability of a power system with respect to a preset assumed disturbance and performing necessary control in advance, evaluating dynamic stability with respect to the assumed disturbance by an eigenvalue sensitivity analysis method, The generator output adjustment required to stabilize the eigenvalue determined to be unstable as a result is determined by the sensitivity coefficient of the eigenvalue real part with respect to the generator output and the maximum variation of the eigenvalue real part obtained from the dynamic stability evaluation results. And a preventive control method for the dynamic stability of the power system, which is obtained as a solution of the linear programming method and performs the preventive control of the dynamic stability of the power system.