JP2001025168A - Method of parallel processing in voltage reliability analysis of power system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain voltage stability of each virtual accident and evaluate the reliability on the system by the parallel computation using a plurality of computers, to the virtual accident where a transmission line or a generator is cut off. SOLUTION: This voltage reliability analysis method obtains load allowance at the voltage stability limit point after an expected accident by an approximation method based on continuous power flow computation, to the expected accident case covering all transmission lines or all generators within the system, and evaluates the reliability on the voltage of the system by continuous power flow computation, aiming at an accident case being sorted with this load allowance as the index of voltage stability, when system constitution, each apparatus's impedance, each load capacity, each power source and its capacity, a voltage designated value, a variation of load, etc., are given in the system. Each accident case is distributed to plural arithmetic units, and the voltage reliability is computed in a parallel manner by each arithmetic unit, and the obtained voltage stability in all accident cases is collected and intensified into one arithmetic unit, and moreover accident cases are sorted, in accordance with the voltage stability.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a parallel processing method for evaluating voltage reliability of a power system, which is executed by a control computer of a system control station and a distribution office. Specifically, system conditions such as the system configuration in the power system, the impedance of each device (transmission line, transformer, phase adjustment equipment, etc.), each load capacity, each power supply and its capacity, the specified voltage value, and the amount of load change Given a given power line or generator is disconnected at a given time, the assumed accident case is assumed, and the voltage stability in all assumed accident cases is determined by parallel processing of parallel computers and multiple computers connected by a network. Calculate the degree of voltage stability, sort the most severe cases to the cases where there is room using the voltage stability as an index, select the severe accident case, and perform detailed power flow calculation for voltage reliability evaluation ,
The present invention relates to a parallel processing method in voltage reliability analysis of a power system.

[0002]

2. Description of the Related Art The reliability evaluation of an electric power system is based on the effect of the accident on the system when an accident occurs in the system from the viewpoint of line overload, voltage drop, voltage stability, transient stability, etc. It is performed by the assumed accident analysis to be evaluated. Among them, the evaluation using the voltage stability as an index is called Voltage Security Assessment (VSA). In the conventional voltage reliability evaluation, for each assumed accident case, the power flow calculation is sequentially performed to calculate the load margin value up to the voltage stability limit point (power flow limit point), and the magnitude of this load margin value is determined by the voltage. Assumed accident cases were sorted as direct indicators of stability, and power flow calculations were performed in order of accident cases with the highest severity to evaluate and analyze the voltage reliability of the power system.

[0003]

According to the above-mentioned conventional method, detailed evaluation by power flow calculation is performed for all assumed accident cases, so that the calculation time is enormous. Therefore, application to online control or the like was impossible. Therefore, the present invention enables online high-speed voltage reliability analysis by parallelizing each operation process such as voltage stability calculation, sorting of assumed accident cases according to voltage stability, and detailed evaluation calculation. It is an object of the present invention to provide a parallel processing method.

[0004]

According to the first aspect of the present invention, an approximate calculation method based on a continuous power flow calculation is applied to a voltage stability calculation, and a certain transmission in a system is performed. A case where a power line or a generator is disconnected is regarded as one assumed accident case, and the assumed accident case targeting all transmission lines or generators in the system is distributed to multiple arithmetic units and voltage stability calculation is processed in parallel. And then
High-speed analysis is made possible by sorting a plurality of assumed accident cases by one arithmetic unit according to the voltage stability.

[0005] Further, in the invention according to claim 2, the sorting processing of the voltage stability calculated by each arithmetic unit is parallelized, and all the sorting results are merged, thereby further increasing the speed. Further, according to the third aspect of the present invention, the severe accident cases in the sorted assumed accident cases are subjected to detailed reliability evaluation calculation in parallel, thereby further increasing the speed.

[0006] As described in claims 4 to 6,
In each of the parallel processing described above, by adding a function of adjusting the number of accident cases distributed to all the arithmetic devices to the higher-level arithmetic device, a difference in the arithmetic performance and arithmetic load of each arithmetic device is absorbed, The speedup by parallel processing is not impaired. In addition, in the inventions according to claims 7 to 9, all the arithmetic units have a function of adjusting the number of accident cases, so that the number of accident cases shared by each arithmetic unit is autonomously determined, and further calculation is performed. The load was averaged.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a basic processing procedure of the present invention. First, system conditions (system configuration, impedance of each device (transmission line, transformer, phase adjustment equipment, etc.), load capacity, each power supply and its capacity, specified voltage value, load change amount, etc.)
Is input to the host processor (SA1). Next, a case where one transmission line or generator in the system is disconnected is assumed as an assumed accident case, and each assumed accident case targeting all transmission lines or generators is regarded as an all-parallel arithmetic unit (upper arithmetic unit and (Including lower-level processing units) (S
A2).

Next, voltage stability in each assumed accident case is calculated by parallel processing of all the arithmetic units (SA3). Note that the voltage stability is obtained as a load margin value at the nose point of the PV curve. Here, the PV curve is a load (power demand) P indicating how the bus voltage changes with respect to a change in power demand.
And a bus line voltage V. The nose point refers to a voltage stability limit point (power flow limit point).
The load amount corresponding to the difference between the current load and the voltage stability limit point in FIG. 5 described later corresponds to the load margin value. The present invention also has a feature in the approximate calculation method of the load margin value, the details of which will be described later.

Returning to FIG. 1, the voltage stability, that is, the load margin value, obtained in parallel by each of the arithmetic units, is collected in the host arithmetic unit or each of the arithmetic units, and the severest possible accident case is determined according to its magnitude. Are sorted in order from (1) to (3) from the accident cases to which there is room. That is, all assumed accident cases are sorted using the voltage stability as a parameter.

[0010] Based on the result of the sorting, the host computing device selects some severe accident cases from a plurality of assumed accident cases (SA5). In the selection, the lower the voltage stability, the greater the severity, so the voltage stability is used as a reference. Then, the selected severe accident cases are distributed to all the arithmetic units including the upper and lower arithmetic units (SA6).

Thereafter, each arithmetic unit performs a continuous power flow calculation to perform a more detailed voltage reliability evaluation (SA7). Here, the outline of continuous power flow calculation is described in No.254 “Continuat
Development of ion Power Flow System "(by Fukuyama, Fujita and Nakanishi).

The PV curve used for the determination of the voltage stability is obtained by increasing the load of the system, obtaining a higher voltage solution and a lower voltage solution by multiroot analysis at discrete operating points, and connecting these solutions. It has been created in the past by several methods. However, in this method, since only discrete tidal current solutions are basically connected, even if there is a possibility of transitioning to another curve between discrete values, it is not actually known, and the reliability as a curve is Absent. PV
As approaching the nose point of the curve, the convergence of the tidal flow calculation becomes more difficult, and it does not converge near the nose point. For this reason, it takes a long time to calculate, and an accurate curve cannot be drawn. Since the power flow calculation is performed only for discrete loads, an accurate curve cannot be drawn when the control of the transformer tap or the like is performed with a load amount between discrete values.

On the other hand, the continuous type power flow calculation solves the inconvenience of the above-described conventional PV curve creation method. By introducing a parameter λ into the equation f (x) = 0, f This is a method of tracing the movement of the equilibrium point of (x, λ) = 0 and thereby creating a PV curve.
The characteristic is that the curve is traced by a mathematical method, so there is a mathematical guarantee as a curve. There is no difficulty in convergence near the nose point, and a curve can be drawn at high speed. Basically, since a curve for a load that changes continuously is traced, a PV curve that can accurately follow control of a transformer tap or the like can be drawn. And the like.

[0014] The continuous power flow calculation is described in V. Ajja
rapu, et.al., “The ContinuationPower Flow: A Tool f
or Steady State Voltage Stability Analysis ”(IEEET
rans. on Power Systems, Vol.7, No.1, February 199
2.) and HDChiang, et.al., “CPFLOW: A Practical Too
l for Tracing Power System Steady-StateStationary
Behavior Due to Load and Generation Variations ”
(IEEE Trans.on Power Systems, Vol.10, No.2, May 199
5.) and so on.

In the continuous power flow calculation, the power flow equation can be expressed by Equation 1. Equation 1 can be said to be a one-parameter nonlinear system.

[0016]

F (x, λ) = 0, x∈R ⁿ , f∈R ⁿ , λ∈R (λ∈R ¹ : control parameter for change)

In a power system, a one-parameter system corresponds to, for example, a change in active power load or reactive power load in one load bus. The following method is used to trace the change in the position of the stable equilibrium point due to such a change in the load or power generation. Equation 2 is obtained by differentiating Equation 1 with respect to λ. Note that ∂f / ∂λ in Expression 2 is represented by a transposed matrix of Expression 3.

[0018]

F _x (x, λ) dx / dλ + ∂f / ∂λ = 0

[0019]

３f / ∂λ = [∂f ₁ / ∂λ,..., ∂f _n / ∂λ] ^T

By solving Equation 2 with respect to dx / dλ, Equation 4 is obtained. By integrating Equation 4, a solution curve x (λ) in the section [λ ₀ , λ ₁ ] can be obtained.

[0021]

Dx / dλ = −f _x ⁻¹ (x, λ) ∂f / ∂λ

As long as f _x (x, λ) has an inverse matrix, (x ₀ ,
It is guaranteed by the implicit function theorem that there is only one solution through λ ₀ ). PC (predictor-co
By applying the rrector (predictor-corrector) method,
A PV curve is determined.

Here, the PC method is one of the numerical product decomposition methods of a differential equation, and uses the two points, the current point and the point immediately before the current point, to temporarily determine the next point by a certain predictor (predictor). It is a method of determining points and sequentially correcting them based on another correction formula (corrector). Here is the actual predic
A method of tracing a PV curve using tor and corrector will be described.

The purpose of the predictor is to find an approximate point for the next solution. Specifically, as shown in Equation 5, a first-order polynomial (straight line) passing through the current solution and the previous solution
Is used to predict the next solution. Note that in Equation 5,
h is an appropriate step width. FIG. 2 shows the concept of the predictor.

[0025]

(Equation 5)

After the predictor has generated an approximation (see Equation 6) for the next solution (x ^{i + 1} , λ ^{i + 1} ), the error must be corrected.

[0027]

(Equation 6)

In principle, any effective numerical processing for solving a nonlinear algebraic equation can be used as a corrector. A good predictor is the solution (x ^{i + 1} , λ ^{i + 1} )
, A few convergences are sufficient to achieve the required accuracy. In general, correcto
As r, the NR (Newton-Raphson) method is often used. That is, in the corrector, the NR method is applied to the extended power flow calculation (formula 1) using the estimated value obtained by formula 5 as an initial value, thereby converging on a curve to obtain a solution. FIG. 3 shows the concept of curve tracing by the PC method. Voltage stability is obtained by obtaining a load margin value when the load is continuously changed based on the PV curve thus obtained.

FIG. 4 is a conceptual diagram in which the processing procedure of FIG. 1 is developed for, for example, three parallel processing units. In FIG. 2, a root process (arithmetic device 0) corresponds to a higher-level arithmetic device, a process 1 (arithmetic device 1) and a process 2 (arithmetic device 2) correspond to a lower-order arithmetic device. It shows the flow of processing, calculation results, and evaluation results. The processing in FIG.
2 corresponds to the aggregation processing of the calculation results of the voltage stability by SA3, corresponds to the distribution processing of severe accident cases by SA4 to SA6, and corresponds to the aggregation processing of the evaluation results by SA7. In addition to the input of system conditions (step SA1), the root process (computing device 0) calculates the voltage stability (SA3) as one of the lower-order computing devices 1 and 2 as one of the parallel computing devices. (SA4) and detailed evaluation (SA7).

Hereinafter, as an embodiment of the invention described in claim 1 and claim 2, an approximate calculation method (Look-Ahead method) of a load margin value (voltage stability) at a voltage stability limit point based on continuous power flow calculation. Will be described. (1) Calculation of load margin value by Look-Ahead method In this method, only two operation states, a current operation state and an operation state after a load increase obtained by continuous power flow calculation with a large step width, are used. And use the PV curve as 2
Approximate by the following function.

FIG. 5 shows a PV curve obtained by a continuous power flow calculation with a normal step width in consideration of the operation of the control device and the like, and a PV curve obtained by the Look-Ahead method. In the figure, the current operation state (current time T ₁ )
The load state of the load flow solution X ₁ and P ₁ in. The load P ₂ of the following calculation point (load increase time T ₂₎ determined on the P-V curve in the case of taking the step width increases, obtaining a load flow solution X ₂ by continuous power flow calculation. Here, the time T ₁ and time T ₂
And selects the load node i with the steepest voltage drop (change in the absolute value of the voltage), and calculates the PV curve at this load node i (PV curve by continuous power flow calculation).
Is used to approximate the load margin value. The approximation formula can be described by a quadratic function such as Expression 7. In addition,
In Equation 7, a, b, c are coefficients, and V _i is the bus voltage at the load node i.

[0032]

P = a + bV _i + cV _i ²

Here, (V _i1 , P ₁ ) and (V _i2 , P ₂ )
And the differential value of V ₂ with respect to P ₂ at time T ₂ , Equation 8 is obtained. Incidentally, V _i1 is bus voltage load node i at time T _1, V _i2 is the bus voltage of the load node i at time T _2.

[0034]

(Equation 8)

Equation 9 is obtained by modifying Equation 7, and the load margin value P _Vmax at the nose point (V _i = −b / 2c) is as shown in Equation 10.

[0036]

P = c (V _i + b / 2c) ² + a−b ² / 4c ²

[0037]

[Number 10] ^{_{P Vmax = a-b 2 /}} 4c

Therefore, since V _i1 , V _i2 and their derivatives are known in Equation 8, the coefficients a, b, and c can be calculated, and the load margin value P _VmaX can be calculated from Equation 10. However, since this method is based on the formulation using the absolute value of the voltage V, if the absolute value of the voltage as well as the absolute value of the voltage changes at the node i, the above procedure is used. Is also applied to the angle, so that P
The load margin value _Pθmax on the −θ curve can be obtained. In this case, the load margin value P _max is, P _Vmax and Pshita _MAX
Can be obtained by Expression 11 based on

[0039]

[Number 11] _{P max = (P Vmax + Pθ} max) / 2

(2) The parallel processing method according to the first aspect of the invention As described above, the load margin value (voltage stability) is obtained by using the Look-Ahead method, and the voltage reliability is analyzed. Assuming that one power line or generator is disconnected, assuming an accident case, it is necessary to determine the load margin values at the voltage stability limit point after the assumed accident, and to calculate the voltage stability of each assumed accident case from these values. it can. The first embodiment of the present invention speeds up the calculation of the voltage stability.

FIG. 6 is a flowchart showing the processing of each arithmetic unit in this embodiment. It is assumed that there are three arithmetic units in total. Steps S01 to S04 in FIG.
6 correspond to steps SA1 to SA4, respectively, and steps S13 and S23 in FIG. 6 correspond to step SA3 in FIG.

Here, the number N of assumed accident cases is divided by the number of all the arithmetic units to be distributed (3 in this example), and the number of assumed accident cases shared by each arithmetic unit is equalized. If the number of cases is fractional, the arithmetic units are assigned one by one until the fraction is exhausted, so that the number of cases assigned to each arithmetic unit is made as equal as possible. In FIG. 6, the assumed accident cases 1 to N are distributed to three (1 to I, I + 1 to J, J + 1 to N) (step S02), and all the arithmetic units calculate the voltage stability in parallel for each case. Yes (S0
3, S13, S23). When the calculation in all the arithmetic units is completed, the voltage stability of all the assumed accident cases 1 to N is collected in the root process and sorted (S0).
4). According to this embodiment, a plurality of arithmetic units share the voltage stability calculations of all the assumed accident cases almost equally, and perform the parallel processing, so that the calculation can be speeded up.

(3) The parallel processing method according to the second aspect of the present invention In the second aspect of the present invention, the sorting (S04) of the voltage stability in FIG. As shown in (4), the voltage stability calculated by each arithmetic unit is sorted individually (S04 ', S04').
14, S24), and finally sort results are collected into a root process to perform a merge sort (S14).
05). That is, in the present embodiment, not only the calculation of the voltage stability but also the sorting process thereof is performed in parallel, so that a higher speed can be expected.

Next, a detailed evaluation parallel processing method according to the third aspect of the present invention will be described. Look-Ahe mentioned above
Using the sorted result of the voltage stability obtained by the voltage reliability analysis using the ad method, for example, an accident case in which the voltage stability is equal to or higher than a predetermined threshold The detailed evaluation for these severe accident cases is performed by continuous power flow calculation. Here, the load margin value in the above-described voltage stability calculation (step SA3 in FIG. 1) is obtained as an approximate value, but in the invention of claim 3, a more detailed load margin value, that is, a voltage, is obtained by the continuous power flow calculation. Determine stability and determine voltage reliability.

FIG. 8 shows a processing procedure of the present embodiment. Steps S011 to S013 in FIG. 8 correspond to steps SA5 to SA7 in FIG. 1, respectively, and steps S113 and S213 in FIG. This corresponds to step SA7. Note that step S014 in FIG.
Is a process of outputting the evaluation result of step SA7 in the above.

In the example shown in FIG. 8, the number n of severe accident cases (generally, severe accident case n <the number N of assumed accident cases).
Is divided by the number of all arithmetic devices in the same manner as in the voltage stability calculation, and the number of cases shared by each arithmetic device is equalized. When the number of cases is fractional, the arithmetic units are assigned one by one until the fraction is exhausted, and the number of cases assigned to each arithmetic unit is equalized as much as possible. This idea is similar to the distribution in FIGS.

In FIG. 8, three severe accident cases n (1 to
i, i + 1 to j, j + 1 to n) (step S0
12) For each case, all the arithmetic units perform continuous power flow calculation and perform detailed evaluation on severe accident cases (S013, S113, S213). here,
The detailed evaluation is performed by calculating a load margin value after the transmission line or the generator is disconnected and determining the voltage stability. Thereafter, when the calculations (evaluations) in all the arithmetic units are completed, the evaluation results for all severe accident cases 1 to n are collected in the route process and output (S014). In this embodiment, the analysis processing can be further speeded up by performing parallel processing of detailed evaluation by continuous power flow calculation for severe accident cases distributed to each arithmetic unit.

Next, as an embodiment of the present invention, a method of adjusting the load of the arithmetic unit in the parallel processing according to the present invention will be described. In the parallel processing method according to the first, second, or third aspect of the present invention, the number of assumed accident cases (including the distribution of severe accident cases in FIG. 8) distributed to each arithmetic device is adjusted so that each arithmetic device has The load can be adjusted. This is the case when the processing capabilities of the processing units that perform parallel processing are different from each other, or when the processing loads on the processing units are different because different processing units simultaneously perform different processing. This is effective when both device loads are different, and the processing speed can be increased by adjusting the number of distribution cases.

(4) A load adjusting method according to the processing capacity of the arithmetic unit (distribution method of the number of assumed accident cases) according to the fourth embodiment of the present invention. s _i (i
= 1,2, ..., given n _c) as pre-set number, determines the number of cases to be distributed to each computing device in accordance with the processing capability index s _i. This procedure is described below. The following procedure shows a case where the index becomes larger as the processing capacity becomes higher.

Procedure 1) Assume that the total number of cases is N, the processing capability indices of all the arithmetic units are totaled, and this is S. Procedure 2) Let x be a value obtained by dividing N by S, and let m be the remainder. Procedure 3) Processing performance index s _i (i = 1, 2, 2)
.., N _c ), and the number of distribution cases to each arithmetic unit is x
× s _i . Step 4) The number of distribution cases is increased by 1 in order from the arithmetic unit with the highest processing capability index s _i until the remainder m in step 2 becomes zero. The variables in the above procedure, except for the processing capacity index,
All are calculated as integers.

(5) A load adjusting method according to an arithmetic unit load according to an embodiment of the present invention (a method of distributing the number of assumed accident cases) This method measures the arithmetic unit load of each arithmetic unit at the time of execution. Then, the measurement result is used as an arithmetic unit load index l _i (i = 1,
2,..., N _c ). Then, the number of cases to be distributed to each arithmetic device is determined according to the arithmetic device load index l _i . This procedure is described below. The following procedure shows a case where the arithmetic device load index l _i becomes larger as the arithmetic device load increases.

Procedure 1) The processing unit load in each processing unit is measured, and this is set to l _i (i = 1, 2,..., N _c ). Here, the total number of cases is N. Procedure 2) The arithmetic unit load indices of all the arithmetic units are summed, and this is set to L. Procedure 3) Let x be a value obtained by dividing N by L, and let m be the remainder. Procedure 4) Set c _i = x × l _i and the remainder m in procedure 3 is 0
In until, in order from the high computing device having computing device loads the index l _i, the corresponding number of c _i is incremented by one. Step 5) the number of the transfer case of the computing device to N / c _i.
Note that all variables in the above procedure are calculated as integers except for the calculation device load index.

(6) A load adjusting method according to the processing capacity of the arithmetic unit and the load of the arithmetic unit (distribution method of the number of assumed accident cases) according to the sixth embodiment of the present invention. The number of accident cases
Undecided before distribution, until the calculation for all accident cases is completed, in order from the processing unit that has finished processing,
Distribute the accident cases one by one and repeat. This makes it possible to distribute the number of accident cases in consideration of both the processing capacity of the arithmetic unit and the load of the arithmetic unit, and to adjust the arithmetic unit load such that the calculation time is always shortened even in various environments. This procedure is described below.

Procedure 1) Distribute one accident case to each arithmetic unit. Procedure 2) Each arithmetic unit executes the calculation of the distributed accident case, and notifies the end when the calculation is completed. Step 3) One accident case is distributed to the arithmetic unit that has sent the end notification. Step 4) Steps 2 and 3 are repeatedly executed until the calculation of all the accident cases is completed.

Next, as an embodiment of the invention described in claims 7 to 9, a method of adjusting the load on the arithmetic unit in parallel processing according to the present invention will be described. According to the inventions described in claims 7 to 9, as in the load adjustment method described in claim 4, 5, or 6, the number of accident cases to be distributed to each arithmetic unit is determined not only by the route process but also by the root process. , Are determined at the same time in all arithmetic devices, and the number of accident cases shared by each arithmetic device is obtained. In this method, the processing for determining the number of distribution cases is also performed in parallel, thereby reducing the amount of communication between the arithmetic units and further increasing the processing speed.

That is, in the embodiment of the present invention, all the arithmetic units 0 to 2 including the root process individually adjust the number of accident cases which they share, according to the index representing their arithmetic load. By doing so, each arithmetic unit autonomously adjusts the load. Further, in the embodiment of the invention of claim 8, all the arithmetic units 0 including the root process are included.
2 individually adjusts the number of accident cases shared by itself in accordance with an index representing its own operation load, so that each operation device autonomously adjusts the load. Furthermore, in the embodiment of the ninth aspect of the present invention, all the arithmetic units 0 to 2 including the root process autonomously share the accident cases one by one, and the arithmetic unit which has completed the processing further sequentially sorts the accident cases. The load of each arithmetic device is adjusted by sharing the load one by one.

[0057]

DESCRIPTION OF THE PREFERRED EMBODIMENTS As an embodiment of the present invention, IEEE1
Simulation results when the above-described parallel processing method is applied to each system of 4, 30, 57, 118, and 300 buses will be described. The number of severe accident cases selected from the assumed accident cases is 15 for each system. Detailed evaluation calculation (continuous power flow calculation) is performed for these severe accident cases, and the result is output. For this calculation, two computers (CPU: Pentium II-400MHz)
A workstation cluster connected by a 00Mbps-Ethernet network was used.

FIG. 9 shows a comparison result between the execution time according to the present invention and the execution time according to the conventional sequential processing method. Here, the scale of speedup achieved by the present invention (conventional execution time / execution time according to the present invention) is shown on the vertical axis as the average magnification of the results of five executions. As shown in the figure, it is confirmed that the present invention can perform almost twice as fast processing as the conventional sequential processing method, and in particular, about 1.96 times as fast in a 300-bus system as an actual system. Was done.

[0059]

As described above, according to the present invention, voltage stability calculation, sorting of assumed accident cases, detailed evaluation calculation, etc. are performed for a plurality of assumed accident cases in which a transmission line or a generator is disconnected. By parallelizing each operation processing, a series of assumed accident analysis and voltage reliability evaluation can be speeded up.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a basic processing procedure of the present invention.

FIG. 2 is a conceptual diagram of Predictor.

FIG. 3 is a conceptual diagram of a curve trace by the PC method.

FIG. 4 is a conceptual diagram of parallel processing according to the present invention.

FIG. 5 is a diagram illustrating the concept of the Look-Ahead method.

FIG. 6 is a flowchart showing a processing procedure of each arithmetic processing device in the embodiment of the first invention.

FIG. 7 is a flowchart showing a processing procedure of each arithmetic processing unit according to the embodiment of the second invention.

FIG. 8 is a flowchart showing a processing procedure of each arithmetic processing unit according to the embodiment of the third invention.

FIG. 9 is a diagram showing a simulation result as an example of the present invention.

Claims (9)

[Claims] When a system condition such as a system configuration, each device impedance, each load capacity, each power supply and its capacity, a specified voltage value, and a load change amount in an electric power system is given, a certain transmission in the system is provided. One assumed accident case is when the power line or generator is disconnected, and for the assumed accident case targeting all transmission lines or generators in the system,
The load margin value at the voltage stability limit point after the assumed accident is determined by an approximate calculation method based on continuous power flow calculation, and this load margin value is used as an index of the voltage stability of each assumed accident case, and the load margin is sorted according to this voltage stability. In a voltage reliability analysis of a power system that evaluates the voltage reliability of the system by continuous power flow calculation for accident cases, each assumed accident case is distributed to multiple arithmetic units and the voltage stability is calculated by each arithmetic unit. Voltage reliability of the power system characterized by calculating the voltage stability of all the assumed accident cases in a single arithmetic unit and sorting the assumed accident cases according to the voltage stability. Parallel processing method in analysis. 2. When a system condition such as a system configuration, each equipment impedance, each load capacity, each power supply and its capacity, a specified voltage value, and a load change amount in an electric power system is given, a certain transmission in the system is provided. One assumed accident case is when the power line or generator is disconnected, and for the assumed accident case targeting all transmission lines or generators in the system,
The load margin value at the voltage stability limit point after the assumed accident is determined by an approximate calculation method based on continuous power flow calculation, and this load margin value is used as an index of the voltage stability of each assumed accident case, and the load margin is sorted according to this voltage stability. In a voltage reliability analysis of a power system that evaluates the voltage reliability of the system by continuous power flow calculation for accident cases, each assumed accident case is distributed to multiple arithmetic units and the voltage stability is calculated by each arithmetic unit. Is calculated in parallel, the assumed accident cases are sorted in parallel according to the obtained voltage stability, all the sorting results by each processing unit are aggregated in one processing unit, and the voltage stability is calculated by the processing unit. A parallel processing method in voltage reliability analysis of a power system, characterized by merging and sorting assumed accident cases according to the following. 3. A parallel processing method in voltage reliability analysis of a power system according to claim 1, wherein severe accident cases are selected according to a result of sorting of assumed accident cases, and the severe accident cases are sent to a plurality of arithmetic units. A parallel processing method in voltage reliability analysis of a power system, characterized in that each processing unit distributes and performs detailed voltage reliability evaluation by continuous power flow calculation in parallel. 4. The parallel processing method in the voltage reliability analysis of a power system according to claim 1, wherein one of the plurality of arithmetic units includes all of the arithmetic units including the high-order arithmetic unit. A parallel processing method in voltage reliability analysis of a power system, wherein the number of accident cases distributed to the devices is adjusted according to a preset index indicating the processing capability of each of the processing devices, and the load of each of the processing devices is adjusted. . 5. The parallel processing method in the voltage reliability analysis of a power system according to claim 1, wherein one of the plurality of arithmetic units includes all of the arithmetic units including the high-order arithmetic unit. A parallel processing method in voltage reliability analysis of a power system, wherein the number of accident cases to be distributed to devices is adjusted according to an index representing the operation load of each operation device to adjust the load of each operation device. 6. The parallel processing method in the voltage reliability analysis of a power system according to claim 1, wherein one of the plurality of arithmetic units includes all arithmetic units including the high-order arithmetic unit. A power system for distributing accident cases one by one to devices, and further distributing the accident cases one by one to processing devices that have completed processing, thereby adjusting the load of each processing device. Parallel processing method for voltage reliability analysis of the system. 7. The parallel processing method in voltage reliability analysis of a power system according to claim 1, 2, or 3, wherein the plurality of arithmetic units individually indicate the number of accident cases shared by the plurality of arithmetic units and their own calculation loads. A parallel processing method in voltage reliability analysis of a power system, wherein a load of each arithmetic unit is adjusted by adjusting according to the following. 8. The parallel processing method for voltage reliability analysis of an electric power system according to claim 1, wherein the plurality of arithmetic units individually indicate the number of accident cases shared by the plurality of arithmetic units and the index indicating their own arithmetic load. A parallel processing method in voltage reliability analysis of a power system, wherein a load of each arithmetic unit is adjusted by adjusting according to the following. 9. The parallel processing method for voltage reliability analysis of a power system according to claim 1, wherein a plurality of arithmetic units individually share an accident case one by one, and the processing is completed. A parallel processing method in voltage reliability analysis of a power system, wherein the apparatus further adjusts the load of each arithmetic unit by sequentially sharing the accident cases one by one.