CN115912428A - Method for configuring energy storage device of tough power distribution network based on global sensitivity index - Google Patents
Method for configuring energy storage device of tough power distribution network based on global sensitivity index Download PDFInfo
- Publication number
- CN115912428A CN115912428A CN202211636167.8A CN202211636167A CN115912428A CN 115912428 A CN115912428 A CN 115912428A CN 202211636167 A CN202211636167 A CN 202211636167A CN 115912428 A CN115912428 A CN 115912428A
- Authority
- CN
- China
- Prior art keywords
- output
- distribution network
- energy storage
- power distribution
- renewable energy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
A method and a system for configuring an energy storage device of a flexible power distribution network based on a global sensitivity index are disclosed, wherein a renewable energy unit output probability distribution function covering an extreme weather scene is constructed based on collected basic data of the flexible power distribution network and output historical data of the renewable energy unit, and a probability voltage unbalance calculation model is established; and then, constructing an agent model of the power distribution network through a neural network, and generating a global sensitivity index of the output of each renewable energy source unit to the voltage unbalance of the power distribution network as a basis for configuring an energy storage device, thereby relieving the three-phase voltage unbalance of the power distribution network. According to the method, the output fluctuation of the renewable energy is stabilized by using the energy storage device, so that the voltage imbalance of the power distribution network can be effectively inhibited, the energy storage configuration cost can be reduced, and the energy storage configuration effect can be improved.
Description
Technical Field
The invention relates to a technology in the field of planning of a tough power distribution network, in particular to a tough power distribution network energy storage device configuration method based on global sensitivity indexes.
Background
With the rapid development of renewable energy and distributed power generation technologies, a large-scale single-phase renewable energy power generation device is connected to a power distribution network and is overlapped with the influence of extreme and random weather scenes, so that the problem of unbalanced voltage of the power distribution network is caused. Due to the inherent three-phase asymmetry in the power system, it is not possible to completely eliminate the voltage imbalance of the distribution network. However, due to the harmfulness of the voltage imbalance, it is desirable to suppress the voltage imbalance so that its impact on the grid is as small as possible. The conventional voltage unbalance suppression method considers the voltage unbalance condition caused by load unevenness, and evenly distributes single-phase loads on three-phase lines of a power distribution network so as to suppress the voltage unbalance. For example, a static transfer switch is used to dynamically switch residential loads between different phases, thereby dynamically balancing three-phase loads and suppressing voltage imbalance. In addition to this, the load is balanced by parallel reactive elements, and the variable load as a reactive element can be corrected by thyristor-controlled static VAR compensators, with the disadvantage that harmonics are injected into the grid. With the access of large-scale renewable energy sources to the power distribution network, the importance of inhibiting the voltage unbalance of the power distribution network becomes more and more prominent, and the traditional load balancing method obviously cannot meet the inhibition requirement.
Disclosure of Invention
The invention provides a method and a system for configuring a tough power distribution network energy storage device based on a global sensitivity index, aiming at the defects that the voltage imbalance cannot be inhibited based on a power distribution network proxy model and the sensitivity of the voltage imbalance to the active power of renewable energy sources cannot be calculated in the prior art.
The invention is realized by the following technical scheme:
the invention relates to a configuration method of a tough power distribution network energy storage device based on a global sensitivity index, which comprises the steps of constructing a renewable energy unit output probability distribution function covering an extreme weather scene based on collected tough power distribution network basic data and renewable energy unit output historical data, and establishing a probability voltage unbalance calculation model; and then, constructing an agent model of the power distribution network through a neural network, and generating a global sensitivity index of the output of each renewable energy source unit to the voltage unbalance of the power distribution network as a basis for configuring an energy storage device, thereby relieving the three-phase voltage unbalance of the power distribution network.
The toughness distribution network basic data include: basic operation data such as a topological structure, an operation mode, parameters of lines and power equipment, user loads and the like of the power distribution network.
The output historical data of the renewable energy source unit comprises the following data: the method comprises the following steps of collecting power generation power data of a renewable energy source unit in the running process in the past period, and collecting natural factor data such as wind speed and illumination intensity related to the power generation power.
The output probability distribution function of the renewable energy source unit covering the extreme weather scene is obtained by construction based on historical data of the output of the renewable energy source unit, and the correlation among the outputs of different renewable energy source units is described through a dependent structure, and the method specifically comprises the following steps:
the described dependent structure refers to: functional relations of correlation among variables are described, and the functional relations include a linear correlation coefficient matrix, a normal Copula function, a t Copula function, a Gumbel Copula function, a Clayton Copula function, a Frank Copula function and the like.
The probability voltage unbalance calculation model takes the output of the renewable energy source unit as input and takes the probability voltage unbalance of the power distribution network as output, and the probability voltage unbalance calculation model has probability distribution characteristics.
The neural network generates a proxy model of the probability voltage unbalance calculation model based on the input and the output of the toughness distribution network probability voltage unbalance calculation model as training samples, and specifically comprises the following steps:
step i: generating unit output data according to the renewable energy unit output probability distribution function and the dependent structure relationship, taking the data as input samples, and respectively calculating to obtain corresponding output samples of the distribution network voltage unbalance by using the constructed flexible distribution network probability voltage unbalance calculation model;
step ii: constructing a feedforward neural network model of a single hidden layer, and training the neural network model by taking the input sample and the output sample obtained in the step i as a training sample set;
step iii: and taking the trained neural network model as a proxy model for computing the probability voltage unbalance of the flexible power distribution network, and under a given input sample, rapidly computing a corresponding output sample of the voltage unbalance of the power distribution network based on the constructed proxy model.
The global sensitivity index is calculated and realized based on a variance decomposition theory and a simulation method, and specifically comprises the following steps: when f (x) 1 ,…，x n ) Is at R n Upper system ofThe output function is a calculation function of the probability voltage unbalance degree of the flexible power distribution network, and the input vector is x = (x) 1 ,…，x n ) The output of the renewable energy source unit is obeyed to a continuous probability distribution function p (x) 1 ,…，x n ) Real random variables of (2). Let x = (y, z), where vector
1≤i 1 ，...，i s N is not more than 1, s is not less than 1 and is less than n, and y is a subset of x; the complementary subset of y is->
A global sensitivity index is obtained>
Wherein: d is the system output f (x) 1 ,…，x n ) The variance of (a); (y, z) and (y ', z') are two different random vectors independently generated from the joint probability distribution function p (y, z), respectively; z' and>
for distinguishing between a random vector generated from a joint probability distribution function p (y, z) and a conditional probability distribution->
And generating a random vector.
The global sensitivity index S y And when y only comprises one variable, the solving result of the global sensitivity index is the first-order sensitivity index of the variable.
The energy storage device is configured as follows: the renewable energy source units are arranged in a descending order based on the global sensitivity index, the larger the global sensitivity index value is, the more important the corresponding renewable energy source unit is, and then when the access of a limited number of energy storage devices is considered, the energy storage devices are preferentially configured at the renewable energy source units with the larger global sensitivity index value, so that the three-phase voltage imbalance of the power distribution network is effectively relieved.
The invention relates to a system for realizing the method, which comprises the following steps: the system comprises an output probability distribution unit, a sample generation unit, a model training unit and a configuration generation unit, wherein: the output probability distribution unit calculates the output probability distribution function of the renewable energy source unit according to basic data of the tough power distribution network and output historical data information of the renewable energy source unit, the sample generation unit calculates the probability voltage unbalance degree according to the output probability distribution function of the renewable energy source unit to obtain a small amount of input-output samples for neural network training, the model training unit trains a neural network proxy model according to the small amount of input-output samples to obtain a proxy model for quickly obtaining the probability voltage unbalance degree, and the configuration generation unit calculates the overall sensitivity index of the voltage unbalance degree of the power distribution network according to the large-scale input-output samples obtained by calculating the proxy model to obtain the configuration basis of the energy storage device in the tough power distribution network.
Technical effects
The method calculates the probability voltage unbalance index of the power distribution network based on the agent model, calculates the overall sensitivity index of the voltage unbalance of the power distribution network, and configures the energy storage device based on the overall sensitivity index calculation result to inhibit the voltage unbalance of the power distribution network. Compared with the prior art, the method and the device can obviously improve the calculation efficiency of the probability voltage unbalance index and the voltage unbalance global sensitivity index of the power distribution network, and meanwhile, the energy storage configuration based on the global sensitivity index can improve the application effect of the energy storage device.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic diagram of an IEEE-123 system including a Wind Turbine (WT) and a Photovoltaic (PV) unit;
FIG. 3 is a graph illustrating cumulative probability of voltage imbalance at node 108 of an IEEE-123 system under various conditions;
FIG. 4 is a comparison graph of voltage imbalance suppression effects for energy storage device configuration based on global sensitivity index calculation results.
Detailed Description
As shown in fig. 1, this embodiment relates to a method for configuring an energy storage device of a flexible power distribution network based on a global sensitivity index, which includes:
step S1: and acquiring basic data of the tough power distribution network and historical data of the output of the renewable energy source unit.
Step S2: based on historical data of the output of the renewable energy unit, the uncertain output of the renewable energy unit is extracted, and a probability model of the output of the renewable energy and a correlation coefficient matrix among different outputs are determined. The uncertain output of a wind generating set and a photovoltaic generating set in a renewable energy source set in a power distribution network is related to the environment where the sets are located. Closely related to the output of the renewable energy unit is the wind speed or the light intensity of the environment where the renewable energy unit is located, the wind speed and the light intensity respectively obey Weibull distribution and Beta distribution, and corresponding probability distribution parameters need to be calculated when the probability distribution condition is fitted through historical data. The correlation between variables is characterized by a linear correlation coefficient matrix.
And step S3: establishing a toughness distribution network probability voltage unbalance calculation model, wherein the model takes the output of a renewable energy source unit as input and takes the distribution network probability voltage unbalance as output, and the method specifically comprises the following steps: the uncertain output of the renewable energy source unit is used as the input of a system model, and the output of the system model, namely the voltage unbalance degree of the power distribution network, is calculated based on a simulation method
Wherein: VUF is the voltage unbalance; v 2 Is a negative sequence voltage; v 1 Is a positive sequence voltage, the line voltage V of which passes through a three-phase unbalanced line ab 、V bc 、V ca Or the phase voltage is deduced, i.e. [ beta ]>
Wherein: a =1 & lt 120 °, a 2 ＝1∠240°。
The line voltage of the three-phase unbalanced line is obtained through three-phase load flow calculation of the power distribution network, and for a three-phase voltage unbalanced power system, an active power equation and a reactive power equation of each node are as follows:
The three-phase line power flow equation in the three-phase power flow calculation is as follows:
wherein: the equation coefficient satisfies the equality condition->
To sum up, considering the uncertain output of the renewable energy units in the distribution network, the objective of the calculation of the probability voltage unbalances is to represent the voltage unbalances of the electric power system as a function VUF = f (X) of the uncertain input variables, where: and X is an uncertain input variable and is uncertain output of the distributed renewable energy source. The voltage imbalance solution also has a probabilistic nature due to the uncertainty of X. The simulation method uses a large amount of random sampling data as system input, and solves the probability result output by the system. The sample value input by the system basically reflects the distribution characteristic of the input variable, so the probability voltage unbalance solving result based on the simulation method can well reflect the distribution condition of the system output.
And step S4: based on the input sample and the output sample of the tough distribution network probability voltage unbalance degree calculation model, the agent model of the distribution network probability voltage unbalance degree calculation model is constructed by adopting a neural network model, and the method specifically comprises the following steps:
step S41: generating unit output data according to the renewable energy unit output probability distribution function and the dependent structure relationship obtained in the step S2, taking the data as input samples, and respectively calculating to obtain corresponding output samples of the distribution network voltage unbalance by utilizing the constructed flexible distribution network probability voltage unbalance calculation model;
step S42: and (4) constructing a feedforward neural network model of a single hidden layer, and training the neural network model by taking the input sample and the output sample obtained in the step (S41) as a training sample set.
Step S43: and taking the trained neural network model as a proxy model for computing the probability voltage unbalance of the flexible power distribution network, and under a given input sample, rapidly computing a corresponding output sample of the voltage unbalance of the power distribution network based on the constructed proxy model.
Step S5: and obtaining an output sample under a given input sample based on the constructed proxy model, calculating a global sensitivity index of the output of each renewable energy unit to the voltage unbalance of the power distribution network by using the obtained output sample, and evaluating the influence of the output fluctuation of the renewable energy unit to the voltage unbalance of the power distribution network.
Preferably, in the step S5, the calculation of the global sensitivity index is implemented based on a variance decomposition theory and a simulation method. When f (x) 1 ,…，x n ) Is at R n Above systemThe system output function, the input vector is x = (x) 1 ,…，x n ) Obey a continuous probability distribution function p (x) 1 ,…，x n ) Real random variables of (2). Let x = (y, z), where vector
S is more than or equal to 1 and less than n, and is a subset of x; the complementary subset of y is->
The specific calculation formula of the global sensitivity index is as follows:
Step S6: and configuring an energy storage device according to the global sensitivity index calculation result so as to relieve the unbalance of the three-phase voltage of the power distribution network.
Preferably, in step S6, the energy storage device is configured according to the global sensitivity index calculation result, specifically: and (4) based on the global sensitivity index calculation results of the output of each renewable energy source unit obtained in the step (S5), sorting the importance of the renewable energy source units, wherein the larger the global sensitivity index value is, the more important the corresponding renewable energy source unit is, and further when the access of a limited number of energy storage devices is considered, the energy storage devices are preferentially configured at the renewable energy source units with the larger global sensitivity index value, so that the three-phase voltage imbalance of the power distribution network is effectively relieved.
After energy storage device is configured at the renewable energy unit, in order to effectively stabilize the output fluctuation of the renewable energy unit, the embodiment further performs an optimal stabilizing strategy for the energy storage device, specifically: considering that the renewable energy source unit has output data sampling of T moments in an observation period, and constructing an optimization model considering strategies in two aspects of energy storage power configuration and energy storage energy configuration for the energy storage device in the observation period
In other words, in an observation period, the energy storage device configured by the ith renewable energy unit is considered, and the energy storage configuration strategy mainly meets the requirement that the sum of the variances of the injected power of the nodes at the T moments after the energy storage device is configured is minimum.
The operation constraint conditions of the optimization model comprise:
Through specific practical experiments, the method is implemented on the following IEEE-123 node system: 5 renewable energy sets are connected, wind power sets are connected at nodes 32, 66 and 107, and photovoltaic sets are connected at nodes 73 and 99, as shown in fig. 2. The access phases of the 3 wind power sets are respectively a phase c, a phase c and a phase b, and the access phases of the 2 photovoltaic sets are respectively a phase c and a phase b. The data of the wind speed and the illumination intensity are generated based on a Weibull distribution function and a Beta distribution function respectively on the basis of reasonable parameter setting, and meanwhile, the correlation between the wind speed and the illumination intensity variable is also considered.
Set the total sample size to 7 × 10 5 And according to a calculation formula, 3 groups of variable samples are needed for solving the global sensitivity index of each uncertain factor, wherein the 3 groups of variable samples comprise 2 groups of variable samples independently generated according to joint probability distribution and 1 other group of variable samples generated according to conditional probability distribution. Since 2 sets of variable samples, independently generated from the joint probability distribution, can be used to calculate the global sensitivity index for all uncertainties, 5 sets of renewable energies are calculatedThe global sensitivity index of the output response of the source unit to the output response of the system is that the actually needed variable samples are 7 groups, and the scale of each group of samples is 10 5 . And calculating the voltage unbalance degree of the power distribution network corresponding to all the system input samples by utilizing the neural network. Under the condition of considering the correlation among the variables, the sensitivity coefficient of each random access variable to the output response is obtained based on a formula. The voltage unbalance degree at the node 108 is selected as an output response, and the specific numerical values of the global sensitivity indexes of the renewable energy units to the output response are obtained through calculation and are shown in table 1.
Table 1 global sensitivity index of input variables to voltage imbalance at node 108
The output of the 2 renewable energy units with larger numerical results of the global sensitivity index calculation is processed into random variables, that is, only the uncertainty of the important input variable determined by the global sensitivity index is considered, and the cumulative probability distribution result of the voltage unbalance degree at the node 108 when only the uncertainty of the unimportant input variable determined by the global sensitivity index is considered is compared with that shown in fig. 3. As shown in fig. 3, case1 considers the common influence of all 5 input random variables (WT 1, WT2, WT3, PV1, and PV 2), case2 considers only the influence of 2 input random variables (WT 3 and PV 1) whose global sensitivity index is large, and case3 considers only the influence of 2 input random variables (WT 1 and PV 2) whose global sensitivity index is small.
The comparison of the results of case1, case2 and case3 shows that: the influence of uncertainty of the WT3 and PV1 output on the voltage unbalance at the power distribution network node 108 is closer to the influence of all 5 input random variables (WT 1, WT2, WT3, PV1 and PV 2), which indicates that the global sensitivity index effectively identifies the output of the key renewable energy unit influencing the voltage unbalance of the power distribution network.
In this embodiment, three energy storage configuration conditions are further set to illustrate the effect of the global sensitivity index on guiding the energy storage configuration of the flexible power distribution network. The three energy storage configurations include case4: energy storage is not configured; case5: energy storage devices are arranged at WT3 and PV1 (the global sensitivity index is large); case6: energy storage devices are configured at WT1, WT2 and PV2 (global sensitivity index is small). The results for three different energy storage configuration cases are shown in fig. 4.
As shown in fig. 4, after the energy storage device is configured at the renewable energy access node (WT 3 and PV 1) with a large global sensitivity index in case5, the fluctuation of the voltage imbalance degree is significantly reduced; and after the renewable energy source access nodes (WT 1, WT2 and PV 2) with smaller global sensitivity indexes in the case6 are configured with energy storage devices, the fluctuation of the voltage unbalance degree is smaller. The configuration guidance effect of the global sensitivity index on the energy storage device in the tough power distribution network is shown, the energy storage is configured at the renewable energy access node with the larger global sensitivity index, the fluctuation of the voltage unbalance of the power distribution network can be reduced to a greater extent, and the situation that the voltage unbalance of the tough power distribution network exceeds the specified safety standard due to the too large fluctuation is avoided.
Compared with the prior art, the method has the advantages that the influence of the global fluctuation of the renewable energy sources is considered by the energy storage device configuration points obtained by the global sensitivity indexes, the global sensitivity indexes of each renewable energy source unit in the power distribution network on the voltage unbalance of the power distribution network can be quickly and efficiently calculated by using the proxy model, so that the energy storage is preferentially configured at the unit which has the large influence on the voltage unbalance of the power distribution network, the output fluctuation of the renewable energy sources is effectively stabilized, the voltage unbalance of the power distribution network is restrained, and the configuration effect of the energy storage device is fully exerted.
The foregoing embodiments may be modified in many different ways by those skilled in the art without departing from the spirit and scope of the invention, which is defined by the appended claims and all changes that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.
Claims (10)
1. A method for configuring an energy storage device of a flexible power distribution network based on a global sensitivity index is characterized in that a renewable energy unit output probability distribution function covering an extreme weather scene is constructed based on collected basic data of the flexible power distribution network and output historical data of the renewable energy unit, and a probability voltage unbalance degree calculation model is established; then, an agent model of the power distribution network is constructed through a neural network and is used for generating a global sensitivity index of the output of each renewable energy source unit to the voltage unbalance of the power distribution network as a basis for configuring an energy storage device, so that the three-phase voltage unbalance of the power distribution network is relieved;
the toughness distribution network basic data include: the topological structure, the operation mode, the parameters of lines and power equipment and the user load of the power distribution network;
the output historical data of the renewable energy source unit comprises the following data: the method comprises the following steps that generated power data of a period of time in the past, which are collected in the running process of a renewable energy unit, and wind speed and illumination intensity natural factor data related to the generated power;
the probability voltage unbalance calculation model takes the output of the renewable energy source unit as input and takes the probability voltage unbalance of the power distribution network as output, and the probability voltage unbalance calculation model has probability distribution characteristics.
2. The method for configuring an energy storage device of a flexible power distribution network based on a global sensitivity index as claimed in claim 1, wherein the neural network generates a proxy model of the probabilistic voltage imbalance calculation model based on input and output of the probabilistic voltage imbalance calculation model of the flexible power distribution network as training samples, and specifically comprises:
step i: generating unit output data according to the renewable energy unit output probability distribution function and the dependent structure relationship, taking the data as input samples, and respectively calculating to obtain corresponding output samples of the distribution network voltage unbalance by using the constructed flexible distribution network probability voltage unbalance calculation model;
step ii: constructing a feedforward neural network model of a single hidden layer, and training the neural network model by taking the input sample and the output sample obtained in the step i as a training sample set;
step iii: and taking the trained neural network model as a proxy model for computing the probability voltage unbalance of the flexible power distribution network, and under a given input sample, rapidly computing a corresponding output sample of the voltage unbalance of the power distribution network based on the constructed proxy model.
3. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in claim 1, wherein the global sensitivity index is calculated based on a variance decomposition theory and a simulation method, and specifically comprises the following steps: when f (x) 1 ，...，x n ) Is at R n The output function of the system is a calculation function of the probability voltage unbalance degree of the flexible power distribution network, and the input vector is x = (x) 1 ，...，x n ) The output of the renewable energy source unit is obeyed to a continuous probability distribution function p (x) 1 ，...，x n ) Real random variables of (2); let x = (y, z), where vector y = (x) i1 ，…，x is )，1≤i 1 ，...，i s N is less than or equal to n, s is more than or equal to 1 and less than n, and y is a subset of x; the complementary subset of y is
A global sensitivity index is obtained>
Wherein: d is the system output f (x) 1 ，...，x n ) The variance of (a); (y, z) and (y ', z') are two different random vectors independently generated from the joint probability distribution function p (y, z), respectively; z' and>
to distinguish between a random vector generated from a joint probability distribution function p (y, z) and a conditional probability distribution->
A generated random vector.
4. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in claim 1, wherein the configuring the energy storage device is: the renewable energy source units are arranged in a descending order based on the global sensitivity index, the larger the global sensitivity index value is, the more important the corresponding renewable energy source unit is, and then when the access of a limited number of energy storage devices is considered, the energy storage devices are preferentially configured at the renewable energy source units with the larger global sensitivity index value, so that the three-phase voltage imbalance of the power distribution network is effectively relieved.
5. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in any one of claims 1 to 4, wherein the method specifically comprises the following steps:
step S1: acquiring basic data of a tough power distribution network and historical data of output of a renewable energy source unit;
step S2: extracting uncertain output of the renewable energy unit based on historical output data of the renewable energy unit, and determining a probability model of the renewable energy output and a correlation coefficient matrix among different outputs; the wind generating set and the photovoltaic generating set in the renewable energy source set in the power distribution network are considered, and uncertain output of the wind generating set and the photovoltaic generating set is related to the environment where the wind generating set and the photovoltaic generating set are located; closely related to the output of the renewable energy source unit is the wind speed or the light intensity of the environment where the renewable energy source unit is located, the wind speed and the light intensity respectively obey Weibull distribution and Beta distribution, and corresponding probability distribution parameters need to be calculated when the probability distribution condition of the renewable energy source unit is fitted through historical data; the correlation between variables is characterized by a linear correlation coefficient matrix;
and step S3: establishing a computing model of the probability voltage unbalance of the tough power distribution network, wherein the model takes the output of a renewable energy source unit as input and takes the probability voltage unbalance of the power distribution network as output, and specifically comprises the following steps: the uncertain output of the renewable energy source unit is used as the input of a system model, and the output of the system model, namely the voltage unbalance degree of the power distribution network, is calculated based on a simulation method
Wherein: VUF is the voltage unbalance; v 2 Is a negative sequence voltage; v 1 Is a positive sequence voltage, the line voltage V of which passes through a three-phase unbalanced line ab 、V bc 、V ca Or the phase voltage is deduced, i.e. [ beta ]>
Wherein: a =1 & lt 120 °, a 2 ＝1∠240°；
And step S4: based on input samples and output samples of a tough distribution network probability voltage unbalance calculation model, a proxy model of the distribution network probability voltage unbalance calculation model is constructed by adopting a neural network model, and the proxy model specifically comprises the following steps:
step S41: generating unit output data according to the renewable energy unit output probability distribution function and the dependent structure relationship obtained in the step S2, taking the data as input samples, and respectively calculating to obtain corresponding output samples of the distribution network voltage unbalance by utilizing the constructed flexible distribution network probability voltage unbalance calculation model;
step S42: constructing a feedforward neural network model of a single hidden layer, and training the neural network model by taking the input sample and the output sample obtained in the step S41 as a training sample set;
step S43: the trained neural network model is used as a proxy model for computing the probability voltage unbalance of the flexible power distribution network, and under a given input sample, the corresponding power distribution network voltage unbalance output sample can be rapidly computed based on the constructed proxy model;
step S5: obtaining output samples under given input samples based on the constructed proxy model, calculating global sensitivity indexes of the output of each renewable energy source unit to the voltage unbalance of the power distribution network by using the obtained output samples, and evaluating the influence of the output fluctuation of the renewable energy source units to the voltage unbalance of the power distribution network;
step S6: and configuring an energy storage device according to the global sensitivity index calculation result so as to relieve the unbalance of the three-phase voltage of the power distribution network.
6. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in claim 5, wherein the line voltages of the three-phase unbalanced lines are obtained through three-phase load flow calculation of the power distribution network, and for a three-phase voltage unbalanced power system, an active power equation and a reactive power equation of each node are as follows:
wherein: the equation coefficient fulfils the condition->
And/or>
Respectively a transconductance matrix and a transadmittance matrix; />
And
are the self-conductance and self-nano matrices, respectively; />
And/or>
Respectively a real part and an imaginary part of the node voltage; subscripts i and j are the number of nodes; a, b and c are three phases of the power system; superscript α and β are phases;
the three-phase line power flow equation in the three-phase power flow calculation is as follows:
wherein: the equation coefficient fulfils the condition->
7. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in claim 5, wherein in the step S5, the calculation of the global sensitivity index is realized based on a variance decomposition theory and a simulation method; when f (x) 1 ，…，x n ) Is at R n The system output function above, the input vector is x = (x) 1 ，...，x n ) Obey a continuous probability distribution function p (x) 1 ，…，x n ) Real random variables of (2); let x = (y, z), where vector
Is a subset of x; the complementary subset of y is->
The specific calculation formula of the global sensitivity index is as follows: />
Wherein: s y The method is a global sensitivity index of a variable y, when y only comprises one variable, the solving result of the formula is a first-order sensitivity index of the variable, the index can be used for quantitatively evaluating the influence degree of the independent action of the variable on the output response of the system, and the influence degree on the output response of the system is larger if the numerical value is larger; d is the system output f (x) 1 ，…，x n ) The variance of (a); (y, z) and (y ', z') are two different random vectors independently generated from the joint probability distribution function p (y, z), respectively; z' and +>
To distinguish between a random vector generated from a joint probability distribution function p (y, z) and a conditional probability distribution->
A generated random vector.
8. The method for configuring energy storage devices of a flexible power distribution network based on global sensitivity index as claimed in claim 5, wherein in step S6, the energy storage devices are configured according to the calculation result of the global sensitivity index, specifically: based on the global sensitivity index calculation results of the output of each renewable energy source unit obtained in the step S5, importance ranking is carried out on the renewable energy source units, the larger the global sensitivity index value is, the more important the corresponding renewable energy source unit is, and further when the access of a limited number of energy storage devices is considered, the energy storage devices are preferentially configured at the renewable energy source units with the larger global sensitivity index value, so that the three-phase voltage imbalance of the power distribution network is effectively relieved;
after energy storage device is configured at the renewable energy source unit, in order to effectively stabilize output fluctuation of the renewable energy source unit, an optimal stabilizing strategy for the energy storage device is further performed, specifically: considering that the renewable energy source unit has output data sampling of T moments in an observation period, and constructing an optimization model considering strategies in two aspects of energy storage power configuration and energy storage energy configuration for the energy storage device in the observation period
In other words, in an observation period, the energy storage device configured by the ith renewable energy unit is considered, and the energy storage configuration strategy mainly meets the requirement that the sum of the variances of the injected power of the nodes at the T moments after the energy storage device is configured is minimum.
9. The method for configuring the energy storage device of the flexible power distribution network based on the global sensitivity index as claimed in claim 1, wherein the operation constraint conditions of the optimization model include:
10. A system for implementing the global sensitivity index-based method for configuring energy storage devices in a flexible power distribution network according to any one of claims 1 to 9, the method comprising: the system comprises an output probability distribution unit, a sample generation unit, a model training unit and a configuration generation unit, wherein: the output probability distribution unit calculates the output probability distribution function of the renewable energy unit according to basic data of the tough power distribution network and output historical data information of the renewable energy unit, the sample generation unit calculates the probability voltage unbalance according to the output probability distribution function of the renewable energy unit to obtain a small number of input-output samples for neural network training, the model training unit trains a neural network proxy model according to the small number of input-output samples to obtain a proxy model for quickly obtaining the probability voltage unbalance, and the configuration generation unit calculates the overall sensitivity index of the voltage unbalance of the power distribution network according to the large-scale input-output samples calculated by the proxy model to obtain the configuration basis of the energy storage device in the tough power distribution network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211636167.8A CN115912428A (en) | 2022-12-20 | 2022-12-20 | Method for configuring energy storage device of tough power distribution network based on global sensitivity index |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211636167.8A CN115912428A (en) | 2022-12-20 | 2022-12-20 | Method for configuring energy storage device of tough power distribution network based on global sensitivity index |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115912428A true CN115912428A (en) | 2023-04-04 |
Family
ID=86483182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211636167.8A Pending CN115912428A (en) | 2022-12-20 | 2022-12-20 | Method for configuring energy storage device of tough power distribution network based on global sensitivity index |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115912428A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117313304A (en) * | 2023-05-16 | 2023-12-29 | 上海交通大学 | Gaussian mixture model method for analyzing overall sensitivity of power flow of power distribution network |
-
2022
- 2022-12-20 CN CN202211636167.8A patent/CN115912428A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117313304A (en) * | 2023-05-16 | 2023-12-29 | 上海交通大学 | Gaussian mixture model method for analyzing overall sensitivity of power flow of power distribution network |
| CN117313304B (en) * | 2023-05-16 | 2024-03-08 | 上海交通大学 | Gaussian mixture model method for analyzing overall sensitivity of power flow of power distribution network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pradhan et al. | Coordinated power management and control of standalone PV-hybrid system with modified IWO-based MPPT | |
| Wu et al. | Distribution network reconfiguration for loss reduction and voltage stability with random fuzzy uncertainties of renewable energy generation and load | |
| Prakash et al. | Optimal placement and sizing of DG for power loss minimization and VSI improvement using bat algorithm | |
| Bakeer et al. | A sophisticated modeling approach for photovoltaic systems in load frequency control | |
| Moawwad et al. | Dynamic security-constrained automatic generation control (AGC) of integrated AC/DC power networks | |
| CN115912428A (en) | Method for configuring energy storage device of tough power distribution network based on global sensitivity index | |
| Belbachir et al. | Multi-objective optimal renewable distributed generator integration in distribution systems using grasshopper optimization algorithm considering overcurrent relay indices | |
| Abou El-Ela et al. | Multiobjective fruit fly optimization algorithm for OPF solution in power system | |
| Yu et al. | Inverse time overcurrent optimization of distribution network with DG based on improved gray wolf algorithm | |
| Hsu et al. | Artificial neural network based adaptive load shedding for an industrial cogeneration facility | |
| Anjaiah et al. | A real-time DC faults diagnosis in a DC ring microgrid by using derivative current based optimal weighted broad learning system | |
| Satapathy et al. | A mutated hybrid firefly approach to mitigate dynamic oscillations of second order PLL based PV-battery system for microgrid applications | |
| Wang et al. | Robust voltage control model for active distribution network considering PVs and loads uncertainties | |
| Larouci et al. | Cuckoo search algorithm for solving economic power dispatch problem with consideration of FACTS devices | |
| Dilshad et al. | NeuroFuzzy wavelet based auxiliary damping controls for STATCOM | |
| Arefi et al. | A novel algorithm based on honey bee mating optimization for distribution harmonic state estimation including distributed generators | |
| Nath et al. | Design and performance analysis of adaptive neuro fuzzy controller for load frequency control of multi-power system | |
| Aryanezhad et al. | Voltage dip mitigation in wind farms by UPQC based on Cuckoo Search Neuro Fuzzy Controller | |
| Griche et al. | Parallel fuzzy logic and PI controller for transient stability and voltage regulation of power system including wind turbine | |
| Vanitha et al. | Solving non-convex economic load dispatch problem by efficient hybrid simulated annealing algorithm | |
| Kumar et al. | Evaluation of DG impacts on distribution networks | |
| Daqaq et al. | Optimal power flow solution including stochastic renewable resources | |
| Sabo et al. | A Survey of the AVOA Metaheuristic Algorithm and its Suitability for Power System Optimization and Damping Controller Design | |
| Rao et al. | ABC and ANN based minimization of power loss for distribution system stability | |
| Prakash | Multi Sensitivity Factors and FPA for Loss Reduction and Voltage Profile Improvement by Optimal Placement and Sizing of Capacitor in Mesh Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |