KR20230037336A - System and Method for Estimating Non Linear P-Q Droop Curve Based on Kalman Filter - Google Patents

Abstract

The present invention relates to a device and method for estimating a non-linear P-Q droop curve based on a Kalman filter algorithm, which can minimize voltage fluctuations in a connected power system, which are caused by renewable energy output fluctuations by calculating an active power-reactive power relationship for an effective voltage fluctuation to be 0. The device comprises: a voltage/current measurement unit which measures voltage and current at a connection point of distributed power; an active/reactive power calculation unit which calculates active and reactive power to control reactive power optimized for the amount of active power input without using the measured voltage for control; an active power-voltage relationship estimation unit which estimates an active power-voltage relationship using the calculation results by the active/reactive power calculation unit; a reactive power-voltage relationship estimation unit which estimates a reactive power-voltage relationship using the calculation results by the active/reactive power calculation unit; and an active power-reactive power relationship estimation unit which estimates a nonlinear P-Q droop curve based on the Kalman filter algorithm which does not participate in voltage fluctuations due to load fluctuations or output fluctuations of other distributed power sources, but only offsets voltage fluctuations due to output fluctuations of the distributed power source thereof by using the estimation results by the active power-voltage relationship estimation unit and the reactive power-voltage relationship estimation unit. According to the present invention, controlling reactive power optimized for the amount of active power input can solve excessive reactive power input due to voltage measurement errors and a resultant voltage instability phenomenon.

Description

System and Method for Estimating Non Linear P-Q Droop Curve Based on Kalman Filter

The present invention relates to voltage stability improvement, and specifically, by calculating an active power-reactive power relationship for an effective voltage fluctuation to be zero, it is possible to minimize voltage fluctuations in a connected power system caused by renewable energy output fluctuations. An apparatus and method for estimating a nonlinear P-Q droop curve based on a filter algorithm.

The power system, which is at the core of the eco-friendly energy transition, will experience the impact of the energy transition and diversification at the forefront, and the emergence of new devices that are significantly different from existing power system components, such as battery-based energy storage systems (ESS), is at the beginning of the change. Even at the present stage, it is approaching as a very big challenge for the power system.

For example, unlike existing generators driven by prime movers, most generators using renewable energy are determined by environmental conditions such as weather. Due to this, the intermittent output characteristics of renewable energy generators can cause voltage fluctuations in the connected power system, and the characteristics of renewable energy linked to distributed power sources adjacent to demand points make this problem more serious.

The problem of voltage fluctuations in the distribution grid can be mitigated by operating renewable energy to control the voltage of the point of common coupling (PCC) constantly, but current grid connection regulations give grid voltage control authority only to grid operators. Therefore, voltage control of individual renewable energy is not allowed.

On the other hand, apart from grid connection regulations, direct voltage control of individual renewable energy is dangerous due to control interference between multiple renewable energy sources, so a cautious approach is required.

Until now, various studies have been conducted to overcome the technical difficulties in the electric power system expected in the course of a great transition to eco-friendly energy and to maximize the acceptability of renewable energy.

First, there is a study in which voltage fluctuations are mitigated by updating the Q-V droop curve based on the current analysis of the entire power system. It requires enormous input of operational resources for analysis, so free market entry of small-scale independent power producers (IPPs) cannot be allowed.

In addition, there is a study on the impact of distributed power supply on voltage stability in a three-phase unbalanced distribution system, but it is inappropriate to apply operation as a technology that can be applied in the planning stage.

On the other hand, there is a study that controls the input amount of renewable energy through droop and compares it with conventional synchronous generator control. This uses a preset droop coefficient, so the power system is reanalyzed every time according to the number and location of distributed power sources input. Therefore, it is difficult to apply in reality, such as setting the droop coefficient.

There is also a study that calculates the optimal droop coefficient individually for each distributed power source without depending on the entire operating system. You cannot escape the limits of technology.

Therefore, there is a demand for the development of a new technology capable of minimizing the voltage fluctuation of the connected power system caused by the output fluctuation of renewable energy.

Republic of Korea Patent Publication No. 10-2016-0028751 Republic of Korea Patent No. 10-1763071 Republic of Korea Patent Publication No. 10-2020-0073039

The present invention is to solve the problems of the voltage stability improvement technology of the prior art, and calculates the active power-reactive power relationship for the effective voltage fluctuation to be zero, thereby reducing the voltage fluctuation of the connected power system caused by the output fluctuation of renewable energy. Its purpose is to provide a device and method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm that can be minimized.

The present invention does not use the directly measured voltage for control through the new P-Q droop control, which is different from the existing Q-V droop control, and controls the reactive power optimized for the amount of active power input, thereby controlling excessive reactive power due to voltage measurement error. Its purpose is to provide a device and method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm that can solve input and voltage instability caused by it.

The present invention is not involved in voltage fluctuations due to load fluctuations or output fluctuations of other distributed power sources, and offsets only voltage fluctuations caused by output fluctuations of the distributed power sources themselves, so that the grid operator's authority to control the grid voltage is not violated. An object of the present invention is to provide an apparatus and method for algorithm-based nonlinear P-Q droop curve estimation.

In the present invention, the P-Q droop curve is calculated based on the active/reactive power and voltage measured at the connection point of the distributed power supply, and the nonlinear P-Q droop based on the Kalman filter algorithm enables effective removal of noise inevitably exposed during the measurement process. Its purpose is to provide an apparatus and method for curve estimation.

Other objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned above will be clearly understood by those skilled in the art from the description below.

To achieve the above object, an apparatus for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention includes a voltage/current measurement unit for measuring voltage and current at a linkage point of a distributed power supply; An active/reactive power calculation unit that calculates active and reactive power in order to control reactive power optimized for the input amount of active power without using it; Estimating an active power-voltage relationship using the calculation result in the active/reactive power calculation unit Active power-voltage relationship estimating unit; Reactive power-voltage relationship estimating unit for estimating the reactive power-voltage relationship using the calculation result in the active/reactive power calculation unit; Active power-voltage relationship estimating unit and reactive power-voltage Using the estimation results of the relationship estimation unit, non-linear P-Q droop curve estimation based on the Kalman filter algorithm that cancels only the voltage fluctuations caused by the output fluctuations of the distributed power sources without being involved in the voltage fluctuations caused by the load fluctuations or the output fluctuations of other distributed power sources. It is characterized in that it comprises a; active power-reactive power relationship estimator to.

To achieve another object, a method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention measures voltage and current at a connection point of a distributed power source to improve voltage stability through active/reactive power balance. / Current measurement step; Active/reactive power calculation step of calculating active and reactive power in order to control reactive power optimized for active power input without using the measured voltage for control; Calculation in active/reactive power calculation step An active power-voltage relationship estimating step of estimating an active power-voltage relationship using a result; A reactive power-voltage relationship estimating step of estimating a reactive power-voltage relationship using a calculation result in the active/reactive power calculation step; Using the estimation results of the power-voltage relationship estimation step and the reactive power-voltage relationship estimation step, voltage fluctuations caused by load fluctuations or output fluctuations of other distributed power sources are not involved, and only voltage fluctuations caused by output fluctuations of the distributed power sources are offset. It is characterized by including; an active power-reactive power relationship estimating step of estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm.

As described above, the apparatus and method for nonlinear P-Q droop curve estimation based on the Kalman filter algorithm according to the present invention have the following effects.

First, the active power-reactive power relationship for real voltage fluctuation to be zero is calculated so as to minimize the voltage fluctuation of the connected power system caused by the output fluctuation of renewable energy.

Second, through the new P-Q droop control, which is different from the Q-V droop control, the reactive power control optimized for the active power input amount is performed without using the directly measured voltage for control, resulting in excessive reactive power input due to voltage measurement error and this voltage instability caused by

Third, since voltage fluctuations caused by load fluctuations or output fluctuations of other distributed power sources are not involved, only voltage fluctuations caused by output fluctuations of the distributed power sources are offset, so that the grid operator's authority to control the grid voltage is not violated.

Fourth, the P-Q droop curve is calculated based on the measured active/reactive power and voltage at the connection point of the distributed power supply, and enables effective removal of noise that is inevitably exposed during the measurement process.

1 is a block diagram of an apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm according to the present invention.
2 is a configuration diagram of a 5-bus distribution system linked to a single distributed power source
3 is a block diagram of an independent microgrid system model based on real system data
Figure 4 is a distribution chart of acquisition (blue) and estimation (red) data of parent line 9
5 is a two-dimensional planar projection view of the acquired (blue) and estimated (red) data of FIG. 4;
6 is a graph of the distribution of acquired (blue) and estimated (red) data satisfying the voltage variation range of bus line 9 within ±0.001pu and the active/reactive power curve (black) satisfying the voltage variation 0
7 is a distribution diagram of acquired (blue) and estimated (red) data of parent line 23
FIG. 8 is a two-dimensional planar projection view of the acquired (blue) and estimated (red) data of FIG. 7;
9 is a graph of the distribution of acquired (blue) and estimated (red) data satisfying the bus line 23 voltage fluctuation range within ±0.001pu and the active/reactive power curve (black) satisfying the voltage fluctuation 0
10 is a graph showing improvement in voltage stability by nonlinear droop control in busbars 9 (a, b) and 23 (c, d)
11 is a graph showing improvement in voltage stability by nonlinear droop control of multiple distributed power sources linked to buses 9 and 23;

Hereinafter, a preferred embodiment of a device and method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention will be described in detail.

Features and advantages of the apparatus and method for estimating a nonlinear P-Q droop curve based on the Kalman filter algorithm according to the present invention will become clear through a detailed description of each embodiment below.

1 is a block diagram of an apparatus for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention.

An apparatus and method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention calculates an active power-reactive power relationship so that an effective voltage fluctuation becomes zero, and the voltage of the connected power system generated by the output fluctuation of renewable energy This was done to minimize fluctuations.

In particular, the present invention does not use the directly measured voltage for control through the new P-Q droop control, which is different from the Q-V droop control, and controls the reactive power optimized for the amount of active power input, thereby controlling excessive reactive power due to voltage measurement error. This is to solve the voltage instability phenomenon caused by the input.

To this end, the present invention may include configurations for minimizing the voltage fluctuation of the connection bus according to the intermittent active power output fluctuation of renewable energy.

The present invention may include a process of calculating active power and reactive power through measurement of instantaneous voltage and current, calculating an effective voltage, and constructing distribution data representing a relationship between effective voltage and active/reactive power.

In the present invention, since the distribution data includes measurement noise, noise caused by load fluctuations, and noise caused by fluctuations in peripheral distributed power sources, it may include a configuration for removing them through a Kalman filter algorithm.

Here, the Kalman filter algorithm estimates the relationship between input and output data by removing noise, but the estimation greatly depends on the output data, and the effect of removing noise is not the same as that of the input data. Therefore, since the final product to be obtained in the present invention is the relationship between active power and reactive power, the same noise removal effect is given to active power and reactive power by using effective voltage as an intermediate medium.

To this end, estimation is performed in parallel with two tracks, one for real voltage-active power and the other for real-time voltage-reactive power relationship estimation. Calculate the active power-reactive power relationship to become mathematically.

An apparatus for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention, as shown in FIG. The active/reactive power calculation unit 20 that calculates active and reactive power and the calculation result of the active/reactive power calculation unit 20 in order to control reactive power optimized for the input amount of active power without using it for control Active power-voltage relationship estimation unit 30 for estimating the relationship between active power and voltage, and reactive power-voltage for estimating the relationship between reactive power and voltage using calculation results of the active/reactive power calculation unit 20 Using the estimation results of the relationship estimation unit 40, the active power-voltage relationship estimation unit 30, and the reactive power-voltage relationship estimation unit 40, voltage fluctuations caused by load fluctuations or output fluctuations of other distributed power sources are involved. and an active power-reactive power relationship estimator 50 for estimating a non-linear P-Q droop curve based on a Kalman filter algorithm that cancels only voltage fluctuations caused by output fluctuations of the distributed power supply itself.

The apparatus for nonlinear P-Q droop curve estimation based on the Kalman filter algorithm according to the present invention having such a configuration performs the following steps to improve voltage stability through active/reactive power balance.

A method for estimating a nonlinear P-Q droop curve based on a Kalman filter algorithm according to the present invention includes a voltage/current measurement step of measuring voltage and current at a connection point of a distributed power supply. In order to control the reactive power optimized for the amount of active power input without using the measured voltage for control, the active/reactive power calculation step of calculating active and reactive power and the calculation result in the active/reactive power calculation step are used to an active power-voltage relationship estimating step of estimating an active power-voltage relationship; Reactive power-voltage relationship estimation step of estimating the reactive power-voltage relationship using the calculation result in the active/reactive power calculation step, and using the estimated results of the active power-voltage relationship estimation step and the reactive power-voltage relationship estimation step Active power-reactive power relationship that estimates the nonlinear P-Q droop curve based on the Kalman filter algorithm that offsets only the voltage fluctuations caused by the output fluctuations of the distributed power sources without being involved in the voltage fluctuations caused by the load fluctuations or the output fluctuations of other distributed power sources. Include an estimation step.

A detailed description of the voltage stability improvement through active/reactive power balance is as follows.

2 is a configuration diagram of a 5-bus distribution system linked to a single distributed power source.

Due to the intermittent output characteristics of renewable energy, distributed generation (DG) based on this can cause rapid voltage fluctuations in the connection bus.

2 is a 5-bus distribution system in which distribution lines are simplified, and the voltage decreases due to the load of each bus as it moves from bus 1 to bus 5 before power input of distributed power.

)투입은 해당 모선의 유효전력 부하(

)와 상쇄되어 순 부하를 감소시키는데, 이는 변전소에서 모선 5로 흐르는 전류의 크기를 감소시켜 전 모선의 전압이 상승하는 결과로 이어진다.Active power of the distributed power source connected to bus 5 (

) Input is the active power load of the bus (

) cancels out the net load, which reduces the magnitude of the current flowing from the substation to bus 5, resulting in an increase in the voltage across all busses.

That is, the intermittent output of renewable energy causes voltage fluctuations in the distribution system.

In a general distribution system where the inductance component dominates, consumption of reactive power is linked to a decrease in voltage, so voltage fluctuations caused by intermittent output of renewable energy can be suppressed through consumption of reactive power corresponding to the amount of active power input of distributed power sources. .

The relationship between the voltage and phase of active/reactive power in a steady state is as shown in Equations 1 and 2.

및

는 각각 모선

에 투입되는 유효전력과 무효 전력을 의미하고,

는 모선

및

와 관련된 어드미턴스 행렬 성분을 의미한다.here,

and

are each mothership

Means active power and reactive power input to

is the mothership

and

It means the admittance matrix component related to

및

는 각각 모선

및

의 전압을,

및

는 해당 모선의 위상을 의미하며,

는

의 위상각을 의미한다.also,

and

are each mothership

and

the voltage of

and

denotes the phase of the mothership,

Is

means the phase angle of

The relationship between active/reactive power and voltage at a specific operating point is expressed by Equations 3 and 4, which are derivatives of the voltage of Equations 1 and 2.

의 값을 갖는다.Since the impedance of a line includes a dominant inductance component and a slight resistance component, the phase of the admittance is

has a value of

을 항상 만족하므로, 분산전원의 유효전력 출력 증가는 연계점 전압의 증가로 이어진다.Therefore, in a situation where active power is supplied from distributed power sources,

is always satisfied, an increase in the active power output of the distributed power source leads to an increase in the connection point voltage.

을 항상 만족하며, 분산전원의 무효전력 출력 증가는 연계점 전압의 증가로 이어진다.Similarly, in a situation where reactive power is supplied from a distributed power source,

is always satisfied, and an increase in the reactive power output of the distributed power source leads to an increase in the link point voltage.

는 매우 큰 값을 갖는다. 또한, 인덕턴스 성분의 우세로 인해

는 0°보다 -90°에 근접하므로,

는 -1에 근접한 값을 갖는다. 즉,

의 영향이 상대적으로 크므로, 분산전원이 무효전력을 소량 소모(즉,

)하더라도

의 관계가 성립한다.On the other hand, since the line impedance of the power system is generally very small,

has a very large value. In addition, due to the dominance of the inductance component

is closer to -90° than 0°,

has a value close to -1. in other words,

Since the influence of is relatively large, distributed power consumes a small amount of reactive power (ie,

)even if

relationship is established

Therefore, the voltage increase due to the input of active power of the distributed power source can be offset through appropriate reduction or consumption of reactive power supply.

Meanwhile, the relationship between the voltage of an arbitrary bus line j and the active/reactive power of the bus line i is expressed by Equations 5 and 6.

)되는 것이 일반적이다.When power is supplied from a distributed power source connected to busbar i , it is common for power to flow from busbar i to a nearby busbar j . Therefore, the phase of busbar j lags behind the phase of busbar i (

) is common.

이고.

는 -90°에 근접한다. 따라서,

와

가 성립한다. 이는 주변 모선 j에서의 유효전력 공급량이 증가하거나 부하 소모량이 감소할 때 모선 i의 전압이 감소함을 의미하고, 주변 모선 j에서의 무효전력 소모량이 감소할 때 모선 i의 전압이 증가함을 의미한다.in other words,

ego.

is close to -90°. thus,

and

is achieved This means that the voltage of bus i decreases when the active power supply in the peripheral bus j increases or the load consumption decreases, and the voltage of bus i increases when the reactive power consumption in the peripheral bus j decreases. do.

As a result, the change in active/reactive power in the peripheral bus has a direct effect on the voltage of bus i . In order to offset all of these voltage fluctuations with the distributed power source connected to bus i , excessive reactive power is needed compared to the active power supply capacity of the corresponding distributed power source. capacity may be required.

This corresponds to the case where the distributed power source connected to bus i directly controls the voltage of the connected bus to be constant, and the method based on the PQ droop curve as in the present invention offsets only the voltage fluctuation caused by the active power supply of the distributed power source itself. , fundamentally free from the problem of excessive reactive power capacity.

The detailed description of the nonlinear active/reactive power relationship estimation of the present invention is as follows.

First, the relationship between active and reactive power for maintaining the bus voltage is described as follows.

When the voltage of the distributed power connection point is constant, an increase in reactive power consumption corresponding to an increase in active power supply is defined as in Equation 7.

와

는 각각

와

이다.here,

and

are respectively

and

am.

와

는 계통의 토폴로지에 의해 결정되는 인자로써 유.무효전력의 공급량에는 영향을 받지 않아 상수로 표현된다. 또한, 유.무효전력의 균형 유지에 의해 전압이 일정하게 유지되므로,

도 상수로 표현된다.

and

is a factor determined by the topology of the system and is expressed as a constant, not affected by the amount of active/reactive power supplied. In addition, since the voltage is kept constant by maintaining the balance of active and reactive power,

is also expressed as a constant.

와

는 각각 상수로

는 특정한 유.무효전력 값에서는 항상 동일 값으로 결정된다.thus,

and

are each constant

is always determined to be the same value at a specific active/reactive power value.

That is, since the reactive power consumption that keeps the connected bus voltage constant at a specific active power output is the same as the integration of Equation 7 and is not affected by the operating state of the system, a nonlinear P-Q curve is calculated and applied to the operating state of various systems. can do.

A detailed description of the Kalman filter-based nonlinear P-Q curve estimation is as follows.

Effective error removal of measurement data is important for accurate estimation of the nonlinear P-Q droop curve.

For this purpose, the present invention proposes a P-Q droop curve estimation algorithm based on the Kalman filter algorithm.

The Kalman filter algorithm has excellent smoothing characteristics and strong ability to remove process and measurement noise. In a real environment where the state is accompanied by process noise and measurement errors are always accompanied by observation, the P-Q droop curve can be expressed as a linear time-varying state equation as shown in Equation 8.

,

및 벡터

는 결정 인자이며,

에는 일반적으로 단위행렬

가 사용된다.Here, matrix

,

and vector

is the determinant,

In general, the identity matrix

is used

는 가중치 벡터를 나타낸다.state vector

denotes a weight vector.

는 프로세스 노이즈 벡터, z는 측정된 출력값, 그리고

는 측정 노이즈를 의미한다.

is the process noise vector, z is the measured output value, and

is the measurement noise.

The state vector estimation value is updated by the following process.

Measurement-based update obtains the measured value z ( t ) as follows and computes the values a posteriori.

는 칼만게인,

는 양의 정부호 대칭 행렬, 그리고 r은 특이행렬을 회피하기 위해 임의로 선택된 양수이다.here,

is the Kalmangain,

is a symmetric positive definite matrix, and r is an arbitrarily chosen positive number to avoid singular matrices.

로 결정하며, 여기서 I는 단위행렬이다.usually,

, where I is the identity matrix.

The time series update is done as follows.

는 양의 정부호 공분산 행렬이고, 본 발명의 일 실시 예에서는 프로세스 노이즈와 측정 노이즈가 상호 독립적이므로 0으로 설정한다.here,

is a positive definite covariance matrix, and is set to 0 because process noise and measurement noise are mutually independent in an embodiment of the present invention.

The time series operation increments time t and repeats the above measurement-based update and time series update.

는 수학식 11에 의해 계산된다.Estimated output after performing the operation

is calculated by Equation 11.

The P-Q curve is estimated based on the measurement of the voltage at the bus line connected to the distributed power supply and the active/reactive power output of the distributed power supply.

In this process, matrices c( t ) and z ( t ) are determined by Equations 12 and 13, respectively.

는 원래의 전압으로부터 분산전원 연계로 인해 변한 전압의 편차를 의미한다. 측정값을 기반으로 수학식 9 및 10의 칼만필터 알고리즘을 통해 추정된 상태

는 수학식 14에서와 같다.here,

Means the deviation of the voltage changed due to the distributed power connection from the original voltage. State estimated through the Kalman filter algorithm of Equations 9 and 10 based on the measured values

Is the same as in Equation 14.

Therefore, the deviation of the estimated voltage is determined as in Equation 15

From this, the reactive power to eliminate the voltage deviation is determined by Equation 16 arranged by setting the left side of Equation 15 to 0.

는 외부요인에 의한 영향을 반영하는 것으로, 주변의 다른 분산전원 변동 또는 부하 변동 등에 의한 연계 모선의 전압 변동을 반영한다.here,

reflects the influence of external factors, and reflects the voltage change of the connected bus due to other distributed power fluctuations or load fluctuations in the vicinity.

That is, it represents a DC component that is not removed by the high-frequency noise removal capability, which is the basic performance of the Kalman filter algorithm.

성분을 제거한 수학식 17에 의해 무효전력을 소모하면, 외부요인에 의한 전압 변동에는 반응하지 않고 분산전원 자체의 유효전력 공급에 의한 전압 변동만 상쇄할 수 있다.On the other hand, when the active power supply of the distributed power source is 0, the reactive power consumption must also be 0.

If the reactive power is consumed by Equation 17 with components removed, it is possible to offset only the voltage fluctuation caused by the active power supply of the distributed power source itself without reacting to the voltage fluctuation caused by external factors.

As a result, reactive power consumption according to active power supply is immediately determined by a simple algebraic equation without time delay for system analysis.

In order to verify the device and method for estimating the nonlinear P-Q droop curve based on the Kalman filter algorithm according to the present invention described above, the independent microgrid system EMT model of FIG. 3 was constructed based on real system data of domestic island regions.

3 is a configuration diagram of an independent microgrid system model based on real system data.

The distributed generation model is linked to busses 9 and 23, respectively, and assumes a situation in which mutual cooperative control is impossible because they belong to different individual generators.

The diesel generators on Bus 1 are owned by the microgrid operator, which monopolizes the microgrid voltage control rights and duties.

Therefore, the distributed power sources associated with buses 9 and 23 can only adjust their own active/reactive power output.

First, voltage acquisition by random active/reactive power output and nonlinear active/reactive power droop curve estimation will be described.

Figure 4 is a distribution diagram of acquired (blue) and estimated (red) data for parent line 9.

Since there is no mutual information exchange between the distributed power sources connected to buses 9 and 23, the output change of the distributed power source connected to bus 9 cannot be distinguished from the load change from the perspective of the distributed power source connected to bus 23.

Therefore, in order to estimate the relationship between active and reactive power of distributed power connected to bus 9, it is necessary to effectively remove both the load fluctuation and the output change of distributed power connected to bus 23. The relationship between the voltage of bus 9 acquired in various system operation situations and the active/reactive power output of the linked distributed power source is shown in the distribution of blue dots in FIG. 4 .

Here, since the voltage of bus 9 is greatly influenced not only by the output of the distributed power supply connected to the corresponding bus, but also by the output of the distributed power connected to the load and bus 23, FIG. 4 shows the voltage of bus 9 and the .The reactive power output relationship is not clearly shown and contains a lot of noise. The noise can be effectively removed through the Kalman filter algorithm-based estimation method described above, and as a result, as shown in the red dot distribution in FIG. is clearly estimated.

FIG. 5 is a two-dimensional planar projection view of the acquired (blue) and estimated (red) data of FIG. 4 .

5 shows voltage fluctuations according to active and reactive power, and the PQV estimation curve of FIG. and projected on the plane of the vertical axis, respectively.

Measurement data contains a lot of noise at a level that makes it very difficult to estimate trends in all projection planes, so it is very difficult to calculate an effective active/reactive power relationship with conventional linear estimation methods. However, when the PQV curve is projected onto the active power-PQV estimation curve and the vertical axis, and the reactive power-PQV estimation curve and the vertical axis, the estimation results are excellent with almost no noise.

The estimated value in (c) of FIG. 5 expresses a relationship with active power as a single curve without noise according to active power. Here, since it is a curve rather than a straight line, when looking from one end of the line to the other end, the distribution of voltage direction is observed due to curvature, but it can be inferred that this is not noise.

In (d) of FIG. 5 , it appears as a thick line including the voltage direction distribution due to the influence of these bends, but this is not noise.

After extracting data whose voltage fluctuation according to active and reactive power output is within ±0.001 pu from the PQV estimation curve of FIG.

6 is a graph of the distribution of acquired (blue) and estimated (red) data that satisfies the voltage fluctuation range of bus line 9 within ±0.001pu and the active/reactive power curve (black) that satisfies the voltage fluctuation of 0.

Since the distribution of blue dots includes the influence of load fluctuations and output fluctuations of bus 23, a single reactive power corresponding to a specific active power cannot be calculated.

On the other hand, it can be confirmed that the distribution of red dots that has been estimated has a narrow distribution width of reactive power corresponding to a specific active power.

The red dot is calculated by the estimated polynomial. When the voltage fluctuation is perfectly zero, the distribution width of the red dot converges to zero.

That is, a single reactive power for making the voltage fluctuation due to input of specific active power to zero is simply calculated. The red dot distribution in FIG. 6 appears together as the range of voltage fluctuation is widened to within ±0.001 pu because it is extremely difficult to graph the distribution of blue dots under the condition that the voltage fluctuation is perfectly zero.

This can be confirmed by the fact that the solid black line in FIG. 6 showing the estimated polynomial curve crosses the center of the red dot distribution. The black solid line is a line expressed based on Equation 16, and shows that a significant amount of reactive power must be supplied when the active power is zero. Since this is only a value for offsetting the voltage caused by external factors, the final P-Q droop expressed by Equation 17 is obtained by moving the black line in FIG. pass through the point

The relationship between the voltage of the bus 23 acquired in various system operation situations and the active/reactive power output of the linked distributed power source is shown in the distribution of blue dots in FIG. 7 . In the same way as in the case of bus line 9, the relationship between the voltage of bus line 23 and the active/reactive power output of the linked distributed power supply is clearly estimated as one PQV curve, as shown in the red dot distribution in FIG.

8 shows voltage fluctuations according to active and reactive power, and the PQV estimation curve of FIG. and projected on the plane of the vertical axis, respectively.

Measurement data contains a lot of noise at a level that makes it very difficult to estimate trends in all projection planes, so it is very difficult to calculate an effective active/reactive power relationship with conventional linear estimation methods. However, when the PQV curve is projected onto the active power-PQV estimation curve and the vertical axis, and the reactive power-PQV estimation curve and the vertical axis, the estimation results are excellent with almost no noise.

The estimated value in (c) of FIG. 8 expresses a relationship with active power as a single curve without noise according to active power. Here, since it is a curve rather than a straight line, when looking from one end of the line to the other end, the voltage direction distribution is observed due to curvature, but it can be inferred that this is not noise.

In (d) of FIG. 8 , it appears as a thick straight line including the distribution of the voltage direction due to the influence of these bends, but this is not noise.

After extracting data whose voltage fluctuation according to active and reactive power output is within ±0.001 pu from the PQV estimation curve of FIG.

Since the distribution of blue dots includes the influence of load fluctuations and output fluctuations of Bus 9, a single reactive power corresponding to a specific active power cannot be calculated. On the other hand, it can be confirmed that the distribution of red dots that has been estimated has a narrow distribution width of reactive power corresponding to a specific active power.

10 is a graph showing improvement in voltage stability by nonlinear droop control at busbars 9 (a, b) and 23 (c, d).

The red dot is calculated by the estimated polynomial, and when the voltage fluctuation is perfectly zero, the distribution width of the red dot converges to zero.

That is, a single reactive power for making the voltage fluctuation due to input of specific active power zero is simply calculated. The red dot distribution in FIG. 9 is shown together as the range of voltage fluctuation is widened to within ± 0.001 pu because it is extremely difficult to graph the distribution of blue dots under the condition that the voltage fluctuation is perfectly zero. This can be confirmed by the fact that the solid black line in FIG. 10 showing the estimated polynomial curve crosses the center of the red dot distribution.

The black solid line is a line expressed based on Equation 16, and shows that a significant amount of reactive power must be supplied when the active power is zero. Since this is only a value for offsetting the voltage caused by external factors, the final P−Q droop expressed by Equation 17 is obtained by moving the black line in FIG. pass through the point

The distributed power source operated based on the active/reactive power droop curve estimated above can maintain the voltage fluctuation of the connected bus line due to power supply to the grid as zero.

10 shows voltage fluctuations of bus lines 9 and 23 according to the amount of active power supplied between the distributed power source operated by the method according to the present invention and the existing distributed power source operated with a power factor of 1.

When the distributed power supply of bus 9 supplies 0.5 MW (distributed power supply of bus 23 is stopped), it can be confirmed that the operation based on the nonlinear droop curve improves voltage stability by maintaining the highest bus voltage below 1.05 pu. Similarly, when the distributed power supply of bus 23 supplies 0.5 MW (distributed power supply of bus 9 is stopped), the operation based on the nonlinear droop curve reduces the maximum bus voltage from 1.065 pu to 1.05 pu, confirming that the voltage stability is improved.

11 is a graph showing improvement in voltage stability by nonlinear droop control of multiple distributed power sources connected to buses 9 and 23.

Even when operating multiple distributed power sources, voltage stability can be improved through the proposed nonlinear droop curve-based operation. When the distributed power sources of buses 9 and 23 each supply 0.25 MW for a total of 0.5 MW, it can be confirmed that the operation based on the nonlinear droop curve improves voltage stability by reducing the maximum bus voltage from 1.06 pu to 1.05 pu as shown in Figure 10. .

That is, as shown in FIG. 11 showing the voltage according to the active power when the distributed power supply is operated with reactive power 0 (ie, power factor 1) and when operated using the estimation result (left), the reactive power is In case of 0, the voltage is between 1.06 and 1.045 and between 1.04 and 1.025, whereas when using the proposed method, it can be seen that the voltage fluctuation width is effectively reduced between 1.055 and 1.045 and between 1.03 and 1.025.

The apparatus and method for estimating the nonlinear P-Q droop curve based on the Kalman filter algorithm according to the present invention described above converges the bus voltage fluctuation by the distributed power supply operation to 0 through the distributed power supply operation based on the PQV curve estimation using the Kalman filter. Through this, the voltage fluctuation applied to the entire microgrid was also minimized.

Therefore, it is possible to input a very high level of renewable energy-based distributed power.

As described above, it will be understood that the present invention is implemented in a modified form without departing from the essential characteristics of the present invention.

Therefore, the specified embodiments should be considered from an explanatory point of view rather than a limiting point of view, and the scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the equivalent range are considered to be included in the present invention. will have to be interpreted

10. Voltage/current measuring part
20. Active/reactive power calculation unit
30. Active power-voltage relationship estimation unit
40. Reactive power-voltage relationship estimation unit
50. Active power-reactive power relationship estimation unit

Claims (16)

a voltage/current measurement unit for measuring voltage and current at a linkage point of distributed power;
an active/reactive power calculation unit that calculates active and reactive power in order to control reactive power optimized for active power input without using the measured voltage for control;
an active power-voltage relationship estimator for estimating an active power-voltage relationship using a calculation result of the active/reactive power calculator;
a reactive power-voltage relationship estimator for estimating a reactive power-voltage relationship using the calculation result of the active/reactive power calculation unit;
Using the estimation results of the active power-voltage relationship estimator and the reactive power-voltage relationship estimator, voltage fluctuations caused by load fluctuations or output fluctuations of other distributed power sources are not involved, and only voltage fluctuations caused by output fluctuations of the distributed power sources are offset. An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, comprising: an active power-reactive power relationship estimator for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm. The method of claim 1, wherein the relationship between the voltage and phase of active/reactive power in a steady state is

is defined as,
here,

and

are each mothership

Means active power and reactive power input to

is the mothership

and

Means the admittance matrix component related to ,

and

are each mothership

and

the voltage of

and

denotes the phase of the mothership,

Is

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that it means a phase angle of The method of claim 2, wherein the relationship between active and reactive power and voltage at a specific operating point is

is defined as,
Since the impedance of a line includes a dominant inductance component and a slight resistance component, the phase of the admittance is

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that it has a value of The method of claim 3, in a situation in which active power is supplied from a distributed power source

is always satisfied, an increase in the active power output of the distributed power source leads to an increase in the connection point voltage,
In a situation where reactive power is supplied from a distributed power source,

Apparatus for estimating a nonlinear PQ droop curve based on the Kalman filter algorithm, characterized in that always satisfying, and an increase in the reactive power output of the distributed power source leads to an increase in the linkage point voltage. 4. The method of claim 3, wherein the relationship between the voltage of a given bus line j and the active/reactive power of a bus line i is:

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that defined as The method of claim 1, wherein in the active power-reactive power relationship estimation unit,
When the voltage of the distributed power connection point is constant, the increase in reactive power consumption corresponding to the increase in active power supply is,

is defined as,
here,

and

are respectively

and

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that. According to claim 6,

and

is a factor determined by the topology of the system and is expressed as a constant, not affected by the amount of active/reactive power supplied,
Since the voltage is kept constant by maintaining the balance of active and reactive power,

is also expressed as a constant,

and

are each constant

Apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that is always determined to be the same value at a specific active/reactive power value. The method of claim 6, wherein in the active power-reactive power relationship estimation unit,
In a real environment where measurement errors are always accompanied, the PQ droop curve is

is defined as,
Here, matrix

,

and vector

is the determinant,

In general, the identity matrix

is used, and the state vector

is the weight vector,

is the process noise vector, z is the measured output value, and

Apparatus for nonlinear PQ droop curve estimation based on the Kalman filter algorithm, characterized in that means measurement noise. 9. The method of claim 8, wherein the measurement-based update of the state vector estimate comprises:

Obtain the measured value z ( t ) with , compute the values a posteriori,
here,

is the Kalmangain,

is a symmetric positive definite matrix, and r is a positive number randomly chosen to avoid singular matrices,

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that I is an identity matrix. The method of claim 9, wherein the time series update of the state vector estimate is

made up of,
here,

Is a positive definite covariance matrix, and is set to 0 if process noise and measurement noise are mutually independent. 11. The method of claim 10, wherein the time series operation increments time t and repeats the above measurement-based update and time series update,
Estimated output after performing the operation

Is

is calculated as,
A device for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that the PQ curve is estimated based on the measurement of the voltage and the active and reactive power output of the distributed power supply in the distributed power supply line. The method of claim 11, wherein the matrices c ( t ) and z ( t ) are, respectively,

is determined by
here,

Apparatus for estimating a nonlinear PQ droop curve based on the Kalman filter algorithm, characterized in that means the deviation of the voltage changed due to the distributed power supply linkage from the original voltage. The method of claim 11, the state estimated through the Kalman filter algorithm based on the measured value

Is,

An apparatus for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, characterized in that defined as The method of claim 13, wherein the deviation of the estimated voltage is

, and the reactive power to eliminate the voltage deviation is arranged by setting the left side to 0,

is determined by
here,

Apparatus for estimating the nonlinear PQ droop curve based on the Kalman filter algorithm, which reflects the influence of external factors, and reflects the voltage fluctuation of the connected bus due to other distributed power fluctuations or load fluctuations. 15. The method of claim 14,

ingredients removed

Device for nonlinear PQ droop curve estimation based on Kalman filter algorithm, characterized in that when reactive power is consumed by, it does not react to voltage fluctuations caused by external factors and cancels only voltage fluctuations caused by active power supply of the distributed power source itself. To improve voltage stability through balance of active and reactive power,
A voltage/current measurement step of measuring voltage and current at a connection point of a distributed power source;
An active/reactive power calculation step of calculating active and reactive power in order to control reactive power optimized for active power input without using the measured voltage for control;
an active power-voltage relationship estimation step of estimating an active power-voltage relationship using a calculation result in the active/reactive power calculation step;
a reactive power-voltage relationship estimation step of estimating a reactive power-voltage relationship using a calculation result in the active/reactive power calculation step;
Using the estimation results of the steps of estimating active power-voltage relationship and estimating reactive power-voltage relationship, only voltage fluctuations caused by output fluctuations of the distributed power sources are not involved in voltage fluctuations caused by load fluctuations or output fluctuations of other distributed power sources. A method for estimating a nonlinear PQ droop curve based on a Kalman filter algorithm, comprising: an active power-reactive power relationship estimating step of estimating a nonlinear PQ droop curve based on an offsetting Kalman filter algorithm.

Images