WO2016026355A1

WO2016026355A1 - Voltage sag simulation and evaluation method of active power distribution grid

Info

Publication number: WO2016026355A1
Application number: PCT/CN2015/083154
Authority: WO
Inventors: 贾东梨; 刘科研; 盛万兴; 胡丽娟; 何开元
Original assignee: 国家电网公司; 中国电力科学研究院; 国网北京市电力公司
Priority date: 2014-08-18
Filing date: 2015-07-02
Publication date: 2016-02-25
Also published as: CN104156892A; CN104156892B

Abstract

A voltage sag simulation and evaluation method of an active power distribution grid, comprising: step 1: constructing a model via an existing power distribution grid analysis software, and calculating a network trend; step 2: generating power distribution grid failure data and distributed power source capacity data via two-point sampling, and determining a simulation solution; step 3: conducting a simulation calculation to calculate an expectation, a variance, a third central moment, a probability density function f(ξ) and a SARFI_x index of a voltage sag of each node; step 4: calculating a probability density function f(ξ)_MC and a SARFI'_x index of a voltage sag amplitude of each node via a Monte Carlo method; step 5: respectively comparing the f(ξ) with the f(ξ)_MC, and the SARFI_x index with the SARFI'_x index, and if a comparison requirement is not satisfied, then returning to step 2. Compared with the prior art, the voltage sag simulation and evaluation method of an active power distribution grid synthetically considers various short-circuit faults in the power distribution grid, thus being suitable for a power distribution network with various wiring modes, and having a high calculation efficiency.

Description

Simulation and evaluation method for voltage drop of active distribution network

Technical field

The invention relates to a voltage drop simulation and evaluation method, in particular to a method for simulating and evaluating a voltage drop of an active distribution network.

Background technique

Voltage Sag, also known as voltage sags or voltage dips, is an event in which the rms voltage of the supply voltage suddenly drops to 90%-10% of the rated voltage amplitude in a short period of time. The typical continuous event is 0.5. -30 weeks. The voltage drop is the most important power quality problem affecting the normal and safe operation of electrical equipment and the normal production of industrial users. The hazards mainly include:

1: affecting the normal use of electricity for residents' work and life;

2: The impact of voltage dips has caused huge economic losses to industrial users;

3: Causes casualties and equipment damage.

With the development of smart distribution networks, more and more distributed power generators (DGs) are connected to the distribution network. When the distributed power supply is connected to the distribution network, the operation state of the distribution network will be changed, and the distribution network will be changed from a passive network to an active network, which will be related to the power quality, power supply reliability and safety of the distribution network. Sexuality, economics, etc. will have an impact. At the same time, more and more sensitive power equipment is connected to the distribution network, and voltage drop has become a power quality problem to be solved in the active distribution network.

Methods for evaluating voltage dips include measured statistical methods and stochastic prediction methods. The measured statistical method requires a long period of time and consumes a lot of money, and the information obtained by the power quality detector has great limitations. Therefore, the reliability of the measured statistical method obtained during the research period is not high. The stochastic prediction method considers the occurrence of a fault as a random probability event. By establishing a system model for the existing fault statistics, the voltage drop caused by the fault is theoretically calculated, thereby effectively detecting the voltage drop and facilitating effective suppression. Measures to reduce voltage, thereby improving the reliability of the power system.

The stochastic prediction method includes the fault point method, the critical distance method, and the Monte Carlo method. The fault point method simulates the characteristics of the entire power system using only a specific fault at several selected points, and the fault is a random process that can occur anywhere in the system, so the specific fault cannot be simulated using several selected points. System trend characteristics. The critical distance method is applicable to radiated networks and is not suitable for active power distribution with grid-like power supply mode. network. The number of samples of the Monte Carlo method is independent of the scale of the system, and the complexity of the system has little effect on it, but the Monte Carlo method has the disadvantages of static, low computational efficiency and long time.

In summary, in order to overcome the above-mentioned drawbacks of the prior art, it is necessary to provide a simulation and evaluation method capable of effectively and quickly determining the voltage drop point of the active distribution network.

Summary of the invention

In order to meet the needs of the prior art, the present invention provides an active distribution network voltage drop simulation and evaluation method, the method comprising the following steps:

Step 1: Construct an active distribution network model with distribution network analysis software, and calculate the voltage values of each node of the active distribution network;

Step 2: Use the two-point estimation method to obtain the distribution network fault data and distributed power capacity data.

Fixed simulation scheme;

Step 3: Simulate the simulation model, calculate the expected value, variance, third-order central moment, probability density function f(ξ) and evaluation index SARFI _x index of the voltage drop amplitude of each node of the active distribution network;

Step 4: Calculate the probability density function f(ξ) _MC and the evaluation index SARFI' _x index of the voltage drop amplitude of each node in the active distribution network by Monte Carlo method;

Step 5: Compare the probability density function f(ξ) with the probability density function f(ξ) _MC , and compare the SARFI _x index with the SARFI′ _x index. If the comparison requirement is not met, return to step 2 to reacquire The simulation scheme.

Preferably, determining the simulation solution in the step 2 includes:

Step 2-1: Determine the random variable X _{i of the} active distribution network according to the distribution network fault data and the distributed power capacity data; and according to the probability density function of each random variable X _i

Calculating the mean μ _{i of} each random variable X _i ; the i=1, 2, . . . , n, n is the dimension of the random variable matrix X;

Step 2-2: Determine a value point on both sides of the mean μ _{i of} a random variable X _i , and use the values x _{i, k of the} two points as the value of the i-th estimated point of the simulation scheme. The value of the random variable is set to the mean value μ _i corresponding to each random variable, k=1, 2;

The position coefficient of the estimated point x _i,k is

The weighting coefficient of the estimated point x _i,k is

Wherein λ _i,k is a k-th order central moment normalized by a random variable X _i , the λ _i,k =E[(X _i -μ _i ) ^k ]/(σ _i ) ^k ;

Preferably, the random variable X _i comprises a line failure rate, fault location, fault type, fault duration, fault impedance, the capacity of wind turbines and photovoltaic power generation system capacity; number of the simulation program is 2 × n;

Preferably, in step 3, the two-point estimation method is used to calculate the expected value and the variance value of the voltage drop amplitude of each node, including:

Step 3-1: construct a nonlinear function Y=h(X) of the voltage drop amplitude of each node based on the random variable X;

Step 3-2: replacing the set of estimated point weight coefficients ω _i,k obtained in the step 2 with the joint probability density of the nonlinear function Y; the constraint condition of the estimated point weight coefficient ω _i,k is

Step 3-3: calculating the voltage drop amplitude of each node is h (μ ₁ , μ ₂ , ..., x _{i, k} , μ _n );

Step 3-4: passing the set of estimated point weight coefficients ω _i,k and the h(μ ₁ , μ ₂ , . . . , x _i,k , μ _n )

To:

The expected value is

The variance value is σ ² = E(Y ² )-[E(Y)] ² ;

Preferably, in the step 3, the Cornith-Fisher series is used to expand the respective central moments of the voltage drop amplitudes of the nodes of the active distribution network and the semi-invariants χ _{i of the} respective orders to obtain the nodes. The probability density function f(ξ) of the voltage drop amplitude is:

among them,

a probability density function that is a standard normal distribution;

Preferably, the SARFI _x index in the step 3 includes SARFI _90% , SARFI _80% , SARFI _70%, and SARFI _50% ;

Preferably, if the compared error value in step 5 is greater than the error threshold, the grid fault data or the distributed power source capacity data is modified to obtain a new simulation scheme.

The superior effects of the present invention compared to the closest prior art are:

1. The method for simulating and evaluating the voltage drop of an active distribution network provided by the present invention comprehensively considers the influence of various short-circuit faults on the voltage drop in the distribution network, and the evaluation result can truly reflect the actual operation process of the distribution network. Medium voltage drop;

2. The method for simulating and evaluating the voltage drop of the active distribution network provided by the invention is applicable not only to the radiation distribution network, but also to the distribution network of the ring network, grid and other wiring modes;

3. The invention provides an active distribution network voltage drop simulation and evaluation method, which can be applied to a distribution network with high permeability distributed power access, and meets the development needs of the smart grid in China;

4. The invention provides a method for simulating and evaluating voltage drop of an active distribution network, and uses a two-point method to convert a random probability problem into multiple deterministic problems, which greatly reduces the number of simulations and significantly improves the calculation efficiency;

5. The invention provides an active distribution network voltage drop simulation and evaluation method, and the result is a quantitative investment analysis, a program comparison, and a reduction of power quality hazard for the power department and the user in the planning and operation stages. Measures are a very necessary scientific basis.

DRAWINGS

The invention will now be further described with reference to the accompanying drawings.

1 is a flow chart of a method for simulating and evaluating a voltage drop of an active distribution network in an embodiment of the present invention.

detailed description

Embodiments of the invention are described in detail below, examples of which are illustrated in the accompanying drawings The same or similar reference numerals indicate the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the drawings are intended to be illustrative of the invention and are not to be construed as limiting.

The invention provides an active distribution network voltage drop simulation and evaluation method by using a two-point method to convert a random probability problem into a plurality of deterministic problems, and constructing a model in an existing distribution network analysis software for active power distribution. The voltage drop simulation of the network, statistical analysis of the simulation results, calculation of the statistical characteristics of the voltage drop, analysis of the weak links in the power grid, based on the Cornish-Fisher series to establish the voltage drop probability density function of each node, statistical voltage drop indicators, to achieve The voltage drop simulation and evaluation of the source distribution network provides reference for taking measures to suppress voltage drop and improve the power supply reliability of the distribution network. The specific steps of the simulation and evaluation method for the voltage drop of the active distribution network in this embodiment as shown in FIG. 1 are as follows:

1. Construct the active distribution network model with the existing distribution network analysis software, and calculate the voltage values of each node of the active distribution network; the distribution network analysis software in this embodiment includes the power system simulation software DIgSLIENT, the power system Analysis software Cymedist et al.

2. After obtaining the original fault data of the power grid and the original distributed power source capacity data by using the statistical simulation method, the fault data of the distribution network and the distributed power source capacity data are sampled by the two-point estimation method to obtain a simulation scheme; The statistical simulation method is the Monte Carlo method;

(1) determining a random variable X;

1: Grid fault data includes line failure rate, fault location, fault type, fault duration, and fault impedance;

The line failure rate is uniformly distributed according to [0, 1], and the number of failures per line is proportional to the line failure rate;

The fault location obeys the uniform distribution of [0, 1], that is, the probability of failure at each point on the line is the same;

Fault types include, but are not limited to, single-phase ground short-circuit fault, two-phase ground short-circuit fault, two-phase phase-to-phase short-circuit fault, and three-phase ground short-circuit fault; fault type obeys [0,1] evenly distributed, and each line fails type It is proportional to the probability of occurrence of the fault type;

The fault duration obeys a normal distribution with a expected deviation of 0.06 s and a standard deviation of 0.01 s;

The fault impedance follows a normal distribution with a desired 5 Ω and a standard deviation of 1 Ω.

2: Distributed power capacity data includes capacity data of wind turbines and photovoltaic power generation systems;

A, wind turbine:

In this embodiment, a curve model is used to obtain the relationship between the output power of the wind turbine and the wind speed, that is, the standard power characteristic curve of the wind turbine, and the relationship between the output power of the wind turbine and the wind speed is as follows:

among them:

Both are constants; v _r , p _r are the rated wind speed and rated power of the wind turbine; v _ci , v _co are the cut-in and cut-out wind speeds of the wind turbine.

When the number of wind turbines is N _wtg , the output power of the wind turbine is:

P _ω =P _wind N _wtg (2)

Wherein, the P _wind is the output power of a single wind turbine.

The wind speed probability distribution generally uses the probability density function of the two-parameter Weibull distribution:

Where k is the shape parameter, reflecting the characteristics of the wind speed distribution; c is the scale parameter, reflecting the average wind speed in the region.

When v _ci <p _w <v _r , the wind turbine outputs a probability density function:

B. Photovoltaic power generation system:

The solar photovoltaic system is mainly composed of a solar cell array, a controller and an inverter; the output power of the solar cell array is:

P _solar =rAη (6)

Where r is the irradiance in W/m ² ;

The total area of the solar array and the photoelectric conversion efficiency, respectively, M is the number of battery modules of the solar cell array, and A _m and η _m are the area and photoelectric conversion efficiency of the individual battery components, respectively;

The solar irradiance r can be approximated as a Beta distribution over a certain period of time, and its probability density function is:

Where r _max is the maximum radiance, and α and β are the Beta distribution shape parameters;

The probability density function of P _solar can be obtained from equation (7):

Among them, R _solar = r _max Aη is the maximum output power of the solar cell array; photovoltaic power generation systems generally only provide active power to the grid, and its reactive power can be ignored.

(2) Determine the simulation scheme by two-point estimation method;

1: According to the distribution network fault data and the distributed power capacity data, determine the random variable X of the active distribution network; in this embodiment, the random variable X includes the line failure rate, the fault location, the fault type, the fault duration, and the fault impedance. And the capacity of the wind turbine and the capacity of the photovoltaic system; based on the probability density function of each random variable X _i , calculate the mean of the random variables μ _i , i = 1, 2, 3, 4, 5, 6, 7, random The dimension of the variable matrix X is n=7; therefore the number of simulation schemes is 2×n, ie (x _1,1 ,μ ₂ ,μ ₃ ,...,μ _n ), (x _1,2 ,μ ₂ ,μ ₃ ,...,μ _n ), (μ ₁ ,x _2,1 ,μ ₃ ,...,μ _n ), (μ ₁ ,x _2,2 ,μ ₃ ,...,μ _n ),...,(μ ₁ ,μ ₂ ,...,x _n,1 ), (μ ₁ , μ ₂ ,..., x _n,2 ).

A value point is determined on both sides of the mean μ _i of a fault variable X _i , and the value x _{i,k of the} two points is used as the value of the i-th estimated point in the simulation scheme, and the values of other random variables are set. The mean μ _i corresponding to each random variable, k=1, 2;

Estimate the position coefficient of point x _i,k as

The weight coefficient of the estimated point x _i,k is

Where λ _i,k is the k-th order center distance normalized by the random variable X _i ;

The estimated point x _i,k can be expressed by the mean μ _i and the standard deviation σ _i as:

x _i,k =μ _i +ζ _i,k σ _i

(9)

In addition, the calculation formula of the jth-th order center distance after the random variable X _{i is} normalized is:

λ _ij =E[(X _i -μ _i ) ^j ]/(σ _i ) ^j

(10)

In the formula (10), j=1, 2, ..., 2m-1, in the present embodiment, the two-point estimation rule j=3 is used; the weighting coefficient ω _{i, k} and the position coefficient ξ _{i, k} are defined as follows:

3. Simulate the simulation scheme determined in step 2 with the distribution network analysis software, and calculate the expected value, variance value and center moment of the voltage drop amplitude of each node of the active distribution network by two-point estimation method;

(1) Construct a nonlinear function Y=h(X) of the voltage drop amplitude of each node based on the random variable X; h is an expression of the nonlinear function; in this embodiment, a two-point estimation method is adopted, that is, by matching the first moments of the function h(X), so that m=2 probability sets are used instead of h(X); When the random variable X is an n-dimensional random variable, the point estimation method uses m×n probability sets instead of joint probability density, that is to say, a total of m×n points are used for estimation, and this embodiment adopts 2×7 Point to estimate.

(2) replacing the set of estimated point weight coefficients ω _i,k obtained in step 2 with the joint probability density of the nonlinear function Y; the constraint condition of the estimated point weight coefficient ω _i,k is

(3) Calculate the voltage drop amplitude of each node h (μ ₁ , μ ₂ , ..., x _{i, k} , μ _n ).

(4) By estimating the set of point probabilities ω _i,k and h(μ ₁ , μ ₂ ,..., x _i,k , μ _n ):

1: expected value

The expected value of each moment of the nonlinear function Y is:

2: The variance value is:

σ ² =E(Y ² )-[E(Y)] ²

(12);

4. Calculate the probability density function of the voltage drop amplitude of each node of the active distribution network;

The Cornish-Fisher series is used to expand the various center-to-center distances of the nodes of the active distribution network and the semi-invariants 各_{l of} each order to obtain the probability density function f(ξ) of the voltage drop amplitude of each node;

1: Each order semi-invariant 随机_{l of the} random variable X can be represented by all order origin moments E(X ^l ) not higher than its own order:

The following mathematical relationship exists between the semi-invariants χ _l and the central moments of each order:

2: The Cornish-Fisher series expansion method is an approximation method for obtaining the probability distribution function or probability density function by using the order origin moments of the random variable X and the semi-invariants of each order. The Cornish-Fisher series provides a random variable X probability distribution function whose quantile is a function of the quantile of the standard normal distribution function.

If the mean and variance of the random variable X are μ and σ, respectively, the standard form of the random variable X is ξ=(x-μ)/σ, and the probability density function f(ξ) of the voltage drop amplitude of each node can be expressed as :

among them,

A probability density function that is a standard normal distribution.

3: Obtain a voltage drop evaluation index SARFI _x index according to the voltage drop amplitude of each node of the active distribution network;

The SARFI _x index is used to calculate the probability that the voltage rms value is below the threshold voltage x:

Wherein, N _i is the number of users whose voltage effective value in the study area is lower than the threshold x in the measurement process; N _T is the total number of users in the study area.

In this embodiment, x takes values of ₉₀ , ₈₀ , ₇₀ , and 50 (%), that is, the SARFI _x index includes SARFI _90% , SARFI _80% , SARFI _70%, and SARFI _50% .

5. The probability density function and the voltage drop evaluation index SARFI' _x index of each node of the active distribution network are calculated by Monte Carlo method;

Comparing the probability density function with the probability density function, comparing the voltage drop evaluation index SARFI' _x index with the voltage drop evaluation index SARFI _x index, if the requirements of the voltage drop index of each node are not met, return to step 2 to reacquire the simulation. Program;

In this embodiment, the requirement of the voltage drop index of each node is that the error value of the comparison between the two is less than the error threshold. If the error threshold is 20%, the error value of the comparison is greater than 20%, and then return to step 2 to modify the grid fault data and distribution. Power supply capacity data to obtain a new simulation solution.

Finally, it should be noted that the described embodiments are only a part of the embodiments of the present application, and not all of them. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

Claims

An active distribution network voltage drop simulation and evaluation method, characterized in that the method comprises the following steps:

Step 1: Construct an active distribution network model with distribution network analysis software, and calculate the voltage values of each node of the active distribution network;

Step 2: Using the two-point estimation method to sample the distribution network fault data and the distributed power supply capacity data, and determine the simulation scheme;

Step 3: Simulate the simulation model, calculate the expected value, variance, third-order central moment, probability density function f(ξ) and evaluation index SARFI x index of the voltage drop amplitude of each node of the active distribution network;

Step 4: Calculate the probability density function f(ξ) MC and the evaluation index SARFI' x index of the voltage drop amplitude of each node in the active distribution network by Monte Carlo method;

Step 5: Compare the probability density function f(ξ) with the probability density function f(ξ) MC , and compare the SARFI x index with the SARFI′ x index. If the comparison requirement is not met, return to step 2 to reacquire The simulation scheme.
The method of claim 1, wherein determining the simulation scheme in the step 2 comprises:

Step 2-1: Determine the random variable X i of the active distribution network according to the distribution network fault data and the distributed power capacity data; and according to the probability density function of each random variable X i
Calculating the mean μ i of each random variable X i ; the i=1, 2, . . . , n, n is the dimension of the random variable matrix X;

Step 2-2: Determine a value point on both sides of the mean μ i of a random variable X i , and use the values x i, k of the two points as the value of the i-th estimated point of the simulation scheme. The value of the random variable is set to the mean value μ i corresponding to each random variable, k=1, 2;

The position coefficient of the estimated point x i,k is

The weighting coefficient of the estimated point x i,k is

Wherein λ i,k is a k-th order central moment normalized by a random variable X i , and the λ i,k =E[(X i -μ i ) k ]/(σ i ) k .
The method of claim 1 wherein said random variable X i comprises line failure rate, fault location, fault type, fault duration, fault impedance, wind turbine capacity, and photovoltaic power system capacity; said simulation The number of schemes is 2 x n.
The method according to claim 1, wherein in step 3, the two-point estimation method is used to calculate the expected value and the variance value of the voltage drop amplitude of each node, including:

Step 3-1: construct a nonlinear function Y=h(X) of the voltage drop amplitude of each node based on the random variable X;

Step 3-2: replacing the set of estimated point weight coefficients ω i,k obtained in the step 2 with the joint probability density of the nonlinear function Y; the constraint condition of the estimated point weight coefficient ω i,k is

Step 3-3: calculating the voltage drop amplitude of each node is h (μ 1 , μ 2 , ..., x i, k , μ n );

Step 3-4: obtaining , by the set of the estimated point weight coefficients ω i,k and the h(μ 1 , μ 2 , . . . , x i,k , μ n ):

The expected value is

The variance value is
The method according to claim 1, wherein said step 3 uses a Cornish-Fisher series to measure the respective central moments of the voltage drop amplitudes of the nodes of the active distribution network and the semi-invariants of the respective orders. χ i is expanded, and the probability density function f(ξ) of obtaining the voltage drop amplitude of each node is:

among them,
A probability density function that is a standard normal distribution.
The method of claim 1 wherein said SARFI x index in step 3 comprises SARFI 90% , SARFI 80% , SARFI 70%, and SARFI 50% .
The method according to claim 1, wherein in the step 5, if the compared error value is greater than the error threshold, the grid fault data or the distributed power source capacity data is modified to obtain a new simulation scheme.