Sequential probabilistic power flow power supply capability assessment method, system and device and medium

Abstract

The invention discloses a time sequence probabilistic power flow power supply capability assessment method, system and device and a medium, and relates to the technical field of power supply capability assessment. The method comprises the following steps: firstly, setting a boundary condition of a Gaussian kernel function, constructing a self-adaptive diffusion kernel density estimation model of wind and light output, deriving a diffusion kernel function by adopting a progressive integral mean square error method to select a self-adaptive optimal bandwidth, and improving the local adaptability of the wind and light output model; and secondly, constructing a wind power and photovoltaic time sequence joint probability distribution model by using a Copula function theory, obtaining time sequence wind and light output samples with correlation, and finally performing power supply capability evaluation through a proposed probabilistic power flow power supply capability calculation method considering wind power and photovoltaic time sequence and correlation. The calculation accuracy of the probabilistic power flow power supply capability can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply capacity assessment, and in particular to a method, system, device and medium for assessing power supply capacity of a sequential probabilistic power flow. Background Art

[0002] The deserts, Gobi, and wasteland areas in western China have high-quality wind and solar energy resources, but both wind power generation and photovoltaic power sources have strong randomness and volatility, as well as output timing determined by the distribution of natural resource conditions in the wasteland areas. The system power supply capability (load supply capability, LSC) refers to the maximum load that can be supplied under the constraints of system operation. In order to fully tap the power supply potential of the system and effectively evaluate the power supply capability of the system, the probabilistic flow calculation method is widely used, and considering the timing correlation between wind power and photovoltaic units will seriously affect the probabilistic flow calculation results.

[0003] At present, power scholars have conducted a lot of research on wind and solar power probability flow and power supply capacity assessment methods. Existing technologies mostly use adaptive diffusion kernel density estimation methods to model new energy output, but do not consider the impact of the correlation between wind power and photovoltaic power on the probability flow calculation. At the same time, the timing characteristics between the output of wind and solar units are not considered in the system power supply capacity assessment calculation, which has certain limitations on the reference significance for system scheduling and operation. Summary of the invention

[0004] The purpose of the present invention is to provide a method, system, device and medium for evaluating the time-series probabilistic power flow power supply capacity, which can improve the accuracy of the calculation of the probabilistic power flow power supply capacity.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for evaluating the power supply capability of a sequential probabilistic power flow, comprising:

[0007] Given basic operation information of the power system; the basic operation information includes grid structure information, system operation parameters, load initial power and wind and solar power unit output data;

[0008] Based on the adaptive diffusion kernel density estimation method, the Gaussian kernel function is converted into linear diffusion, the corresponding optimal bandwidth is selected for the diffusion kernel function to achieve self-regulation of the kernel density estimation bandwidth, and the time series probability density models of wind power output and photovoltaic output are respectively established based on the basic operation information;

[0009] Based on the time series probability density model of wind power output and photovoltaic output, the optimal Copula function is selected by using the minimum square Euclidean distance as the correlation measurement and fitting index to construct a time series joint probability distribution model of wind and photovoltaic output;

[0010] Based on the time series joint probability distribution model of wind and photovoltaic output, Gibbs sampling is performed to obtain a joint probability distribution sample group of wind power output and photovoltaic output, and the sample group is substituted into the inverse function of the cumulative probability distribution of wind power output and photovoltaic output to determine the time series related sample group of wind power output and photovoltaic output;

[0011] The time-series related sample group of wind power output and photovoltaic output is input into the probabilistic flow power supply capacity evaluation calculation model, and the system power supply capacity evaluation calculation is performed using the adaptive step-size repeated flow method; wherein the probabilistic flow power supply capacity evaluation calculation model includes a solution model, equality constraints and inequality constraints; the solution model is constructed with the maximum system supply load as the objective function.

[0012] Optionally, the function expression of the adaptive diffusion kernel density estimation method is:

[0013]

[0014] Where f(x; t) is the diffusion kernel density estimate; n is the sample size; K D (·) is the diffusion kernel function; x is a random variable; X m is the measured data of the output of new energy units; t is the time; k is the width of the diffusion core density.

[0015] Optionally, the solution model is expressed as:

[0016]

[0017] Among them, maxP LSC is the maximum load power supplied by the system; N is the number of network nodes; P 0j is the initial load power of node j; λ is the load growth multiple step; P dj is the base power increase of the load at node j.

[0018] Optionally, the equality constraint is expressed as:

[0019]

[0020] Among them, P Gi , Q Gi Respectively represent the active power and reactive power of the generator output at node i; P Li , Q Li are the active power and reactive power of the load at node i respectively; U i , U j Represent the voltage amplitude of nodes i and j respectively; θ ij is the phase angle difference between nodes i and j; G ij , Bij are the conductance and susceptance between nodes i and j respectively.

[0021] Optionally, the inequality constraint is expressed as:

[0022]

[0023] Among them, P Gimax , P Gimin , Q Gimax , Q Gimin Respectively represent the upper and lower limits of the active power output and the upper and lower limits of the reactive power output of the generator at node i; P Gi , Q Gi They represent the active power and reactive power of the generator output at node i respectively; U imax , U imin Respectively represent the upper and lower limits of the voltage at node i; U i represents the voltage at node i; θ ijmax ,θ ijmin Respectively represent the upper and lower limits of the phase angle difference between nodes i and j on both sides of the line; θ ij Indicates the phase angle difference between nodes i and j on both sides of the line; P limax Indicates the upper limit of the active power carried by the branch; N G Represents the set of generator nodes; N l is the power line collection; N B is the set of power system nodes; P l Indicates the active power carried by the branch.

[0024] Optionally, the calculation steps of the adaptive step-size repeated power flow method include:

[0025] Step 1, determine the initial step size λ0 and convergence accuracy ε;

[0026] Step 2: Input the node initial load power P 0j And the growth base power P dj , let P LSC =P 0j ,λ=λ0;

[0027] Step 3: If λ>ε, proceed to the next step; if λ<ε, the calculation ends and outputs P LSC and λ;P LSC is the maximum supplied load power;

[0028] Step 4: Calculate P LSC =P 0j +P dj ;

[0029] Step 5: Use P LSCPerform power flow calculation based on the value of the threshold value to determine whether there is an over-limit. If there is no over-limit, proceed to the next step, otherwise go to step 7.

[0030] Step 6: Let P 0j =P LSC , go to step 4;

[0031] Step 7: The step length is reduced to half of the original value: λ = λ / 2, and then go to step 3;

[0032] Step 8: Compare the maximum supply load power calculated in step 3 with the current load of the system to obtain the system power supply capacity. The calculation formula is as follows:

[0033]

[0034] Where: η is the system power supply capacity; P N The current active load of the system.

[0035] The present invention also provides a sequential probabilistic power flow power supply capability evaluation system, which is applied to the above method and includes:

[0036] A basic parameter unit is set to provide basic operation information of a given power system; the basic operation information includes grid structure information, system operation parameters, load initial power and wind and solar power unit output data;

[0037] A time series probability density model building unit for wind power output and photovoltaic output, used for converting a Gaussian kernel function into a linear diffusion based on an adaptive diffusion kernel density estimation method, selecting a corresponding optimal bandwidth for the diffusion kernel function to achieve self-regulation of the kernel density estimation bandwidth, and establishing time series probability density models for wind power output and photovoltaic output respectively based on the basic operation information;

[0038] A wind-solar output time series joint probability distribution model construction unit is used to select an optimal Copula function based on the time series probability density model of wind power output and photovoltaic output, using the minimum square Euclidean distance as a correlation measure and fitting index to construct a wind-solar output time series joint probability distribution model;

[0039] A sampling unit is used to perform Gibbs sampling based on the time series joint probability distribution model of wind and photovoltaic outputs to obtain a joint probability distribution sample group of wind power output and photovoltaic output, and substitute the sample group into the inverse function of the cumulative probability distribution of wind power output and photovoltaic output to determine a time series related sample group of wind power output and photovoltaic output;

[0040] The system power supply capacity evaluation calculation unit is used to input the time-series related sample group of wind power output and photovoltaic output into the probabilistic power supply capacity evaluation calculation model, and use the adaptive step-size repeated power flow method to perform the system power supply capacity evaluation calculation; wherein, the probabilistic power supply capacity evaluation calculation model includes a solution model, equality constraints and inequality constraints; the solution model is constructed with the maximum system supply load as the objective function.

[0041] The present invention also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the power supply capacity evaluation method according to the above-mentioned sequential probabilistic flow.

[0042] The present invention also provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the method for evaluating the sequential probabilistic power flow power supply capability as described above.

[0043] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0044] The present invention discloses a method, system, device and medium for evaluating the power supply capacity of a sequential probability flow. The method comprises firstly setting the boundary conditions of a Gaussian kernel function, constructing an adaptive diffusion kernel density estimation model of wind and solar power output, using the asymptotic integrated mean square error method to derive the diffusion kernel function to select the adaptive optimal bandwidth, thereby improving the local adaptability of the wind and solar power output model; secondly, using the Copula function theory to construct a sequential joint probability distribution model of wind power and photovoltaic power, and obtaining sequential wind and solar power output samples with correlation, and finally evaluating the power supply capacity by the proposed probabilistic flow power supply capacity calculation method taking into account the sequentiality and correlation of wind power and photovoltaic power. The present invention can improve the accuracy of the calculation of the probabilistic flow power supply capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0046] Figure 1 It is the scheduling plan curve diagram in this embodiment;

[0047] Figure 2 This is a flow chart of the calculation of the time sequence probability power flow power supply capability evaluation in this embodiment;

[0048] Figure 3This is a probability density curve diagram of wind power output in the 9th hour in this embodiment;

[0049] Figure 4 This is the probability density curve of photovoltaic output in the 9th hour in this embodiment;

[0050] Figure 5 This is the probability density curve of wind power output in the 13th hour in this embodiment;

[0051] Figure 6 This is the probability density curve of photovoltaic output in the 13th hour in this embodiment;

[0052] Figure 7 This is a schematic diagram of the IEEE standard 30-node system in this embodiment;

[0053] Figure 8 It is the DC power planning curve diagram in this embodiment;

[0054] Fig. 9 Schematic diagram of load average values ​​in different time periods in this embodiment;

[0055] Fig.10 It is a time series probability density diagram of the power supply capability of the system in this embodiment;

[0056] Fig.11 This is a time series probability density comparison diagram of the system power supply capability in this embodiment;

[0057] Fig.12 is a probability density diagram of the voltage timing of node 4 in this embodiment;

[0058] Fig.13 : is a comparison diagram of the probability density of the voltage timing of node 4 in this embodiment;

[0059] Fig.14 24 is a probability density diagram of the power flow timing of the branch in this embodiment;

[0060] Fig.15 2 is a comparison diagram of the probability density of the power flow timing of branch 24 in this embodiment. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] The purpose of the present invention is to provide a method, system, device and medium for evaluating the time-series probabilistic power flow power supply capacity, which can improve the accuracy of the calculation of the probabilistic power flow power supply capacity.

[0063] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] The present invention provides a method for evaluating the power supply capability of a sequential probabilistic power flow, comprising:

[0065] Firstly, the Gaussian kernel function is transformed into linear diffusion based on the adaptive diffusion kernel density estimation method, and the corresponding optimal bandwidth is selected for the diffusion kernel function to achieve self-regulation of the kernel density estimation bandwidth; secondly, the optimal Copula function is selected by using the minimum square Euclidean distance as the correlation measure and fitting index, and the time series joint probability distribution model of wind and solar power output is constructed, and the time series related samples of wind power and photovoltaic output are extracted; then, taking the maximum system supply load as the objective function, a probabilistic flow power supply capacity evaluation calculation method considering the time series and correlation of wind power and photovoltaic output is proposed; finally, simulation is carried out based on the IEEE30-node system and the measured data of a wind power and photovoltaic power station in Shagohuang area of ​​my country. The results verify that considering the time series and correlation of wind and solar power output can improve the accuracy of the probabilistic flow power supply capacity calculation, and the system power supply capacity calculation results are given in time series, providing more accurate reference information for system uncertainty analysis and scheduling operation.

[0066] A detailed description of the components of each step is as follows:

[0067] 1. Non-parametric estimation method: used to construct diffusion kernel density estimation model.

[0068] First, the kernel density estimation model is constructed: the wind power and photovoltaic output data samples in the Shagohuang area are more volatile, so it is difficult to use a priori parameters or mathematical models to fit and estimate. Therefore, compared with parameter estimation, the direct sample estimation of non-parametric kernel density estimation (KDE) can better reflect the real distribution of data and has a better fitting effect.

[0069] The KDE model f(x) of wind and solar power generation unit output is:

[0070]

[0071] Where: X1, X2, …, X n is the measured output data of new energy units; n is the sample capacity of new energy output data; h is the bandwidth; K(·) is the kernel function. In this paper, Gaussian kernel function is selected as the kernel function, and its expression is:

[0072]

[0073] Combining the above two equations, we get:

[0074]

[0075] When the renewable energy output data sample is given, the bandwidth h has a greater impact on the fitting effect of f(x) than the choice of kernel function. The optimal value of the bandwidth is usually determined by the minimum value of the asymptotic mean integrated square error (AMISE), but under the conventional KDE, the overall optimal bandwidth is still selected, which cannot be automatically adjusted according to the output data density of renewable energy units, and its local adaptability is poor. In order to better fit the uncertainty and volatility of renewable energy unit output, this paper uses the diffusion kernel density estimation method to improve the traditional Gaussian kernel density estimation method and improve its local adaptability.

[0076] Then, the diffusion kernel density estimation model is constructed: In view of the boundary deviation problem caused by the traditional kernel density estimation method in the distribution estimation, Botev sets the initial boundary conditions on this basis and proposes a diffusion kernel density estimation model. This method solves the problem of no probability distribution outside the boundary and ensures the accuracy of the step-by-step estimation value.

[0077]

[0078] Formula (3) is the only solution to the partial differential equation (4). Its boundary conditions are defined as:

[0079]

[0080] Where: δ(xx i ) is x i The Dikla measure at .

[0081] Combining equation (4) and equation (5), we get the adaptive kernel density estimation and its function expression as follows:

[0082]

[0083] Where f(x; t) is the diffusion kernel density estimate; n is the sample size; K D (·) is the diffusion kernel function; x is a random variable; X m is the measured data of the output of new energy units; t is the time; k is the width of the diffusion core density.

[0084] The asymptotic integrated mean square error method is used to derive the diffusion kernel function to select the adaptive optimal bandwidth, ignoring its high-order infinitesimal terms, and the AMISE is obtained as:

[0085]

[0086] The minimum value of AMISE is obtained by derivation of equation (8), and the corresponding bandwidth h* is the optimal bandwidth.

[0087]

[0088] 2. Time Series Joint Probability Distribution Model

[0089] The correlation between the output of wind turbines and photovoltaic units will affect the operating status of the access system. The difference in output between the two in different periods aggravates the impact of their volatility on the system. For example, the photovoltaic output is 0 at night, while wind turbines are generally at their peak output, resulting in the time series difference of wind-solar correlation. A day is divided into 24 periods. According to the probability distribution function of wind-solar output in each period, the time series joint probability distribution model of the two is established using Copula theory, and 24h time series related wind-solar output samples are extracted respectively.

[0090] First, we use the Copula theory to calculate. Copula theory was originally proposed by Sklar in the form of Sklar theorem. It is a measurement method for the dependency function of multivariate extreme value theory. The correlation of n marginal distribution functions can be expressed by the Copula function. Assume F i (x i ,y i ) is the joint cumulative probability distribution function of wind and solar power output in the i-th hour, then:

[0091] F i (x i ,y i )=C i [F 1i (x i ),F 2i (y i )] (10)

[0092] Where: x i ,y i are the wind power output and photovoltaic output data for the ith hour respectively; F 1i (x i ), F 2i (y i ) are the marginal distribution functions of wind power output and photovoltaic output in the ith hour respectively; C i is the connecting Copula function of the two at the ith hour.

[0093] The squared Euclidean distance can reflect the pros and cons of different Copula functions in fitting data. The smaller the value, the better the fitting effect of the function. By comparing the squared Euclidean distances of different Copula functions in each period, the optimal wind-solar joint probability distribution function is selected.

[0094] Then, correlation time series wind and solar power output sampling is performed.

[0095] The optimal time-series wind-solar joint probability distribution function is obtained through the above selection, and then sampling and inverse transformation of the sampled function are performed to obtain time-series wind power output and photovoltaic output samples with correlation. The specific steps are as follows:

[0096] 1) The optimal Copula function is selected using the least square Euclidean distance as the goodness of fit, and the sample group (u ij ,v ij )(j=1,2,…,N).

[0097] 2) Determine the inverse function x of the marginal cumulative probability distribution of wind power output xi and photovoltaic output yi in the i-th hour ij =F 1j -1 (u ij ) and y ij =F 2j -1 (v ij ).

[0098] 3) Substitute the sampled joint probability distribution samples with time series correlation into the inverse function of the marginal cumulative probability distribution of wind power and photovoltaic output respectively to obtain the time series correlation samples of wind power and photovoltaic output.

[0099] 3. Probabilistic power flow power supply capacity evaluation and calculation

[0100] First, a method for realizing the automatic power curve of DC transmission is provided: At present, in most DC transmission projects in my country, a bipolar automatic power control method is adopted in the converter station, and the realization of the automatic power curve is mainly based on the preset power reference value rate sent to the pole control system at time T for automatic execution. Figure 1 It is the power change point and power curve within a certain day.

[0101] At present, the power plan curve issued by the provincial dispatching department is issued at an interval of 15 minutes, and a power plan value is issued as the power reference value of the current period. When two consecutive reference values ​​change, the 15 minutes before the changed power point is used as the power reference value issuance time, and the current changed power point is used as the completion time, and the power adjustment is fully utilized in this 15-minute time period.

[0102] Secondly, construct a probabilistic power flow power supply capacity evaluation calculation model: This embodiment performs a probabilistic power flow power supply capacity evaluation calculation taking into account the correlation between wind power and photovoltaic output, and analyzes the operating status of the power system. The probabilistic power flow power supply capacity evaluation calculation model consists of three parts: solution model, equality constraint conditions and inequality constraint conditions. They are:

[0103] 1) Solving the model

[0104]

[0105] Where: P LSC The maximum load power supplied by the system; N is the number of network nodes; P 0j is the initial load power of node j; λ is the load growth multiple step; P dj is the base power increase of the load at node j.

[0106] 2) Equality constraints

[0107]

[0108] Where: P Gi , Q Gi Respectively represent the active power and reactive power of the generator output at node i; P Li , Q Li are the active power and reactive power of the load at node i respectively; U i , U j Represent the voltage amplitude of nodes i and j respectively; θ ij is the phase angle difference between nodes i and j; G ij , B ij are the conductance and susceptance between nodes i and j respectively.

[0109] 3) Inequality constraints

[0110]

[0111] Where: P Gimax , P Gimin , Q Gimax , Q Gimin Respectively represent the upper and lower limits of the active power output and the upper and lower limits of the reactive power output of the generator at node i; U imax , U imin Respectively represent the upper and lower limits of the voltage at node i; θ ijmax ,θ ijmin Respectively represent the upper and lower limits of the phase angle difference between nodes i and j on both sides of the line; P limax Indicates the upper limit of the active power carried by the branch; N G Represents the set of generator nodes; N l is the power line collection; N Bis the set of power system nodes. On this basis, the adaptive step-size repeated power flow method is used to calculate the time-series probability optimal power flow.

[0112] Then, an adaptive step-size repetitive power flow method is constructed: The selection of the load growth multiple step value is very important in solving the maximum power supply capacity. If the value is too large, the accuracy of the calculation result will be reduced, while if the value is too small, the convergence speed of the calculation will be affected. Therefore, this paper adopts the adaptive step-size repetitive power flow method for solution. The detailed steps are as follows:

[0113] 1) Determine the initial step size λ0 and the convergence accuracy ε.

[0114] 2) Input node initial load power P 0j And the growth base power P dj Let P LSC =P 0j , λ=λ0.

[0115] 3) If λ>ε, proceed to the next step; if λ<ε, the calculation ends and returns to P LSC and λ. LSC That is, the maximum load power that the system can supply under the current operating mode.

[0116] 4) Calculate P LSC =P 0j +P dj .

[0117] 5) P LSC Perform power flow calculation based on the value of the control plane to determine whether there is an over-limit. If there is no over-limit, proceed to the next step; otherwise, go to step 7.

[0118] 6) Let P 0j =P LSC , go to step 4.

[0119] 7) The step size becomes half of the original one, that is, λ=λ / 2, and go to step 3.

[0120] 8) Compare the calculated maximum supply load power with the current system load to obtain the system power supply capacity. The calculation formula is as follows:

[0121]

[0122] Where: η is the system power supply capacity; P N The current active load of the system.

[0123] Finally, the time series probability flow power supply capacity evaluation calculation is performed, such as Figure 2 As shown, the specific steps are:

[0124] 1) Given the basic information of the power system, such as the grid structure information, system operating parameters, initial load power, etc.

[0125] 2) Based on the given wind and solar power unit output data, the adaptive diffusion kernel density estimation method is used to establish the time series probability density models of wind power output and photovoltaic output respectively.

[0126] 3) Based on the Copula theory, the minimum square Euclidean distance is used as the goodness of fit to obtain the optimal time series joint probability distribution model of wind and solar power output.

[0127] 4) Using Gibbs sampling to obtain the joint probability distribution sample group of wind power output and photovoltaic output (u ij ,v ij ).

[0128] 5) will (u ij ,v ij ) is substituted into the inverse function x of the cumulative probability distribution of wind power output and photovoltaic output ij =F 1j -1 (u ij ) and y ij =F 2j -1 (v ij ), obtain the time series related sample group (x ij ,y ij ).

[0129] 6) The wind power output and photovoltaic output time series correlation sample group (x ij ,y ij ) is substituted into the above-mentioned probabilistic power flow power supply capacity evaluation calculation model, and the above-mentioned adaptive step-size repeated power flow method is used to perform system power supply capacity evaluation calculation.

[0130] 4. Example Analysis

[0131] Firstly, the 24-hour measured data of a wind farm and a photovoltaic power station in the Shagohuang area in western my country are used for simulation, and the probability density curve of adaptive diffusion kernel density estimation under time series is compared with the probability density curve of Gaussian kernel density estimation, as shown in Figure 2. Figure 3-Figure 6 As shown (taking 9:00 and 13:00 as examples).

[0132] In the figure, compared with the Gaussian kernel density estimation method, which can only indicate the overall distribution trend of wind power output, the adaptive diffusion kernel density estimation method perfectly describes the peaks of the curves at 2-3MW, 4-6MW, 12-14MW, 19-21MW and 25-27MW, as well as the troughs of the curves at 7-9MW, 17-19MW and 22-24MW. In the figure, the photovoltaic output probability density curve has four peaks at 1-2, 4-6, 9-11 and 17-18, which the Gaussian kernel density estimation method fails to track, thus losing important information on the photovoltaic output probability density. In addition, it can be concluded from the figure that the adaptive diffusion kernel density estimation method not only far exceeds the kernel density estimation method in describing the locations of peaks and troughs, but also is more in line with the probability density of specific output data in terms of the values ​​of peaks and troughs.

[0133] In summary, compared with the traditional Gaussian kernel density estimation method, the adaptive diffusion kernel density estimation curve can achieve better fitting effect for the measured data of wind and solar power output at various times of the day, effectively enhance the local fitting of the wind and solar power output probability density curve, and more accurately reflect the uncertainty and volatility of wind and solar power output.

[0134] Then, this simulation is carried out to simulate the power supply capacity of the time series probability flow. The IEEE 30-node system structure is as follows: Figure 7 As shown. The number of nodes in the network is N B =30, number of branches N l =41, number of generator nodes N G =6.

[0135] Nodes 8 and 21 are selected to connect to wind turbines and photovoltaic units respectively. Both use the adaptive diffusion kernel density probability distribution model, whose data comes from the measured data of wind farms and photovoltaic power stations in Shagohuang area of ​​my country. Node 12 is used as the node for simulating DC transmission power. This simulation does not consider the time interval effect during DC regulation. The DC transmission power plan curve is as follows: Figure 8 As shown. The rest of the load nodes adopt the Gaussian distribution model, and define their load ratings as the steady-state value of the node load. Given the average load of each node in the IEEE 30-node system, Fig. 9 , assuming that the standard deviation of the node timing load is 0.1 of the timing load mean.

[0136] Verify accuracy:

[0137] In order to verify the accuracy and effectiveness of the method of this embodiment, the calculation result of the probability flow power supply capacity at 13:00 is taken as an example, and the calculation result of the Monte Carlo probability flow power supply capacity with a sampling scale of 10,000 times is taken as the accurate result, and the calculation results of the probability flow power supply capacity with and without considering the time series correlation are compared. The comparison of the mean and standard deviation of the node 12, 27 voltage, branch 9, 17 flow, and power supply capacity of the three methods is shown in Table 1.

[0138] Table 1 Calculation results of probability flow power supply capacity

[0139]

[0140] In Table 1, μ V_4 , μ V_27 , μ P_9 , μ P_24 , μ η are the mean values ​​of node 4 and node 27 voltages, branch 9 and branch 24, and power supply capacity respectively; σ V_4 , σ V_27 , σ P_9 , σ P_24 , σ η They are the voltages of node 4 and node 27, branch 9 and branch 24, and the standard deviation of power supply capacity.

[0141] According to the comparison of the table data, the calculation result of the probability flow power supply capacity considering the timing and correlation proposed in this embodiment is relatively accurate, and the error between it and the accurate value is small, which verifies that the calculation result of the probability flow power supply capacity considering the timing correlation of wind and solar power output has high accuracy.

[0142] Calculation results of time-series probability flow power supply capacity: In order to analyze the impact of the timing and correlation characteristics of wind power output and photovoltaic output on the sending-end system, the probability flow power supply capacity is calculated based on the wind and photovoltaic output time-series data at 24:00. Fig.10 The time series probability density diagram of the system power supply capacity shows that the system power supply capacity curve is relatively flat and its peak probability density fluctuates with time. The reason is that in addition to traditional power generation methods at night, wind turbines can make up for the system's output shortage; during the day, the system consists of photovoltaic power generation, traditional power generation methods and a small amount of wind power. The power supply capacity calculation proposed in this paper is to meet the system's safety and stability constraints. The traditional power generation output is adjusted according to the output of wind power and photovoltaic units, resulting in more dispersed system power supply capacity data and a relatively flat curve. In addition, the system's 24h power supply capacity varies between 0 and 60% with time series. The reason is that in addition to meeting the system's DC transmission power demand, it also needs to meet the system's load time series changes as shown in the figure during the day, so the system's power generation will also change accordingly, thereby affecting the power supply capacity.

[0143] Fig.11The figure is a time series probability density comparison diagram of the system power supply capacity, in which the time series characteristic curve takes the wind and solar power unit output at 1:00, 12:00 and 17:00 as input, and the time series characteristic curve without taking into account takes the wind and solar power output of the whole day as input. In the figure, the power supply capacity of the time series characteristic curve without taking into account varies from 20% to 30%, has a peak and the peak probability density is very high, about 0.30, and its data distribution is relatively concentrated, which can only reflect the average trend of the system power supply capacity; while the power supply capacity calculation results of the probability flow power supply capacity of the time series proposed in this paper have greatly expanded the range of power supply capacity at 1:00, 12:00 and 17:00, with multiple peaks and low peak probability density, not exceeding 0.10, and the data distribution is uniform. The power supply capacity between each hour is obviously different, that is, affected by the changes in DC transmission power and load, the power supply capacity at 1:00 at night is strong, while the power supply capacity at 12:00 at noon is relatively small. This shows that the calculation of probabilistic power flow power supply capacity by taking into account the time series characteristics can more accurately describe the time series characteristics of the system power supply capacity and provide more valuable reference information for power system dispatching.

[0144] Fig.12 This is the voltage time series probability density diagram of node 4. As can be seen from the figure, the probability density peak of the node voltage time series probability density diagram fluctuates with the time series change, and the range of node voltage change has obvious differences. The reason is that the time series of wind and solar power output are quite different. When performing probability flow calculation, in order to ensure the maximum power supply capacity of the system, the system node voltage will have time series differences.

[0145] Fig.13 This is a comparison chart of the voltage probability density of node 4. In the figure, the voltage variation range of the node without time sequence characteristics is 0.998~1.002, with the highest peak value of about 2000. The data distribution is concentrated, reflecting the average trend of the voltage of node 4; while the voltage variation range of 1:00, 12:00 and 17:00 of the probability flow calculation results with time sequence is larger, among which 12:00 and 17:00 have two peaks and the probability density is smaller, which is significantly lower than the probability density of the voltage peak of the node without time sequence. The voltage difference between each hour is obvious, that is, affected by the change of DC transmission power and load, the voltage at 1:00 at night is the largest. It can be seen that taking time sequence characteristics into account in the probability flow calculation can accurately describe the fluctuation characteristics of node voltage and provide reliable data for the voltage over-limit analysis of power system nodes.

[0146] Fig.14 This is the branch 24 power flow time series probability density diagram. As can be seen from the figure, the probability density peak of the branch power flow time series probability density diagram also fluctuates with the time series change, and the change range of the branch power flow has obvious differences. This is because when calculating the time series probability power flow, the time series of the solar-wind-solar output data are quite different, which will cause obvious time series differences in the system branch power flow.

[0147] Fig.15 This is a comparison chart of the probability density of branch 24 power flow. As can be seen from the figure, the branch power flow without timing characteristics varies from 5 to 12MW, with a peak value of about 0.4, and the data is concentrated, reflecting the average trend of branch 24 power flow; while the probability power flow calculation results of the time sequence show that the branch 24 power flow at 1:00, 12:00 and 17:00 has a larger variation range, with multiple peaks and a low peak probability density, not exceeding 0.25, and the data is evenly distributed. The power difference between each hour is obvious. It can be obtained that the timing characteristics taking into account the wind and solar power output can more accurately describe the variation range of the branch power flow and provide accurate and reliable timing data reference for line current carrying capacity analysis.

[0148] In summary, the temporal changes in wind and solar power output have a great influence on the probabilistic flow calculation results. Taking into account the temporal characteristics of wind and solar power output in the power supply capacity calculation can give the probabilistic flow power supply capacity calculation results of the system in different time periods, thereby solving the problem of loss of probabilistic characteristics when the temporal sequence is not taken into account, and providing more accurate reference information for power system scheduling and uncertainty analysis.

[0149] Therefore, according to the content of this embodiment, the following beneficial effects are achieved:

[0150] 1) According to the adaptive diffusion kernel density estimation method of wind power and photovoltaic power output, the local adaptability of the output estimation model is improved by utilizing the adaptive characteristics of linear diffusion, which can more accurately describe the uncertainty and volatility of the actual output of wind power stations and photovoltaic power stations.

[0151] 2) Based on the Copula theory, a time series joint probability distribution model of wind power output and photovoltaic output was established with the minimum square Euclidean distance as the evaluation criterion. It was found that wind power output and photovoltaic output in different time periods obey different joint probability distributions, and then the time series related samples of wind power output and photovoltaic output were obtained by sampling.

[0152] 3) A probabilistic power flow capacity calculation method taking into account the timing correlation of wind power output and photovoltaic output is proposed, and it is verified that this method can improve the accuracy of the calculation and give the probabilistic power flow capacity calculation results of the system according to the time period, providing more accurate reference information for system uncertainty analysis and scheduling operation.

[0153] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0154] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for evaluating the power supply capability of a sequential probabilistic power flow, characterized in that: include: Given the basic operating information of the power system; The basic operation information includes grid structure information, system operation parameters, load initial power and wind and solar power unit output data; Based on the adaptive diffusion kernel density estimation method, the Gaussian kernel function is converted into linear diffusion, the corresponding optimal bandwidth is selected for the diffusion kernel function to achieve self-regulation of the kernel density estimation bandwidth, and the time series probability density models of wind power output and photovoltaic output are respectively established based on the basic operation information; Based on the time series probability density model of wind power output and photovoltaic output, the optimal Copula function is selected by using the minimum square Euclidean distance as the correlation measurement and fitting index to construct a time series joint probability distribution model of wind and photovoltaic output; Based on the time series joint probability distribution model of wind and photovoltaic output, Gibbs sampling is performed to obtain a joint probability distribution sample group of wind power output and photovoltaic output, and the sample group is substituted into the inverse function of the cumulative probability distribution of wind power output and photovoltaic output to determine the time series related sample group of wind power output and photovoltaic output; The time-series related sample group of wind power output and photovoltaic output is input into the probabilistic flow power supply capacity evaluation calculation model, and the system power supply capacity evaluation calculation is performed using the adaptive step-size repeated flow method; wherein the probabilistic flow power supply capacity evaluation calculation model includes a solution model, equality constraints and inequality constraints; the solution model is constructed with the maximum system supply load as the objective function.

2. The method for evaluating the power supply capability of sequential probabilistic power flow according to claim 1, characterized in that: The function expression of the adaptive diffusion kernel density estimation method is: Where f(x; t) is the diffusion kernel density estimate; n is the sample size; K D (·) is the diffusion kernel function; x is a random variable; X m is the measured data of the output of new energy units; t is the time; k is the width of the diffusion core density.

3. The method for evaluating the power supply capability of sequential probabilistic power flow according to claim 1, characterized in that: The solution model is expressed as: Among them, maxP LSC is the maximum load power supplied by the system; N is the number of network nodes; P 0j is the initial load power of node j; λ is the load growth multiple step; P dj is the base power increase of the load at node j.

4. The method for evaluating the power supply capability of sequential probabilistic power flow according to claim 1, characterized in that: The equality constraint is expressed as: Among them, P Gi , Q Gi Respectively represent the active power and reactive power of the generator output at node i; P Li , Q Li are the active power and reactive power of the load at node i respectively; U i , U j Represent the voltage amplitude of nodes i and j respectively; θ ij is the phase angle difference between nodes i and j; G ij , B ij are the conductance and susceptance between nodes i and j respectively.

5. The method for evaluating the power supply capability of sequential probabilistic power flow according to claim 1, characterized in that: The inequality constraint is expressed as: Among them, P Gimax , P Gimin , Q Gimax , Q Gimin Respectively represent the upper and lower limits of the active power output and the upper and lower limits of the reactive power output of the generator at node i; P Gi , Q Gi They represent the active power and reactive power of the generator output at node i respectively; U imax , U imin Respectively represent the upper and lower limits of the voltage at node i; U i represents the voltage at node i; θ ijmax ,θ ijmin Respectively represent the upper and lower limits of the phase angle difference between nodes i and j on both sides of the line; θ ij Indicates the phase angle difference between nodes i and j on both sides of the line; P limax Indicates the upper limit of the active power carried by the branch; N G Represents the set of generator nodes; N l is the power line collection; N B is the set of power system nodes; P l Indicates the active power carried by the branch.

6. The method for evaluating the power supply capability of sequential probabilistic power flow according to claim 3 is characterized in that: The calculation steps of the adaptive step-size repeated power flow method include: Step 1, determine the initial step size λ0 and convergence accuracy ε; Step 2: Input the node initial load power P 0j and growth base power P dj , let P LSC =P 0j ,λ=λ0; Step 3: If λ>ε, proceed to the next step; if λ<ε, the calculation ends and outputs P LSC and λ;P LSC is the maximum supplied load power; Step 4: Calculate P LSC =P 0j +P dj ; Step 5: Use P LSC Perform power flow calculation based on the value of the threshold value to determine whether there is an over-limit. If there is no over-limit, proceed to the next step, otherwise go to step 7. Step 6: Let P 0j =P LSC , go to step 4; Step 7: The step length is reduced to half of the original value: λ = λ / 2, and then go to step 3; Step 8: Compare the maximum supply load power calculated in step 3 with the current load of the system to obtain the system power supply capacity. The calculation formula is as follows: Where: η is the system power supply capacity; P N The current active load of the system.

7. A sequential probabilistic power flow power supply capability assessment system, applied to the method according to any one of claims 1 to 6, characterized in that: include: Setting basic parameter units for basic operation information of a given power system; The basic operation information includes grid structure information, system operation parameters, load initial power and wind and solar power unit output data; A time series probability density model building unit for wind power output and photovoltaic output, used for converting a Gaussian kernel function into a linear diffusion based on an adaptive diffusion kernel density estimation method, selecting a corresponding optimal bandwidth for the diffusion kernel function to achieve self-regulation of the kernel density estimation bandwidth, and establishing time series probability density models for wind power output and photovoltaic output respectively based on the basic operation information; A wind-solar output time series joint probability distribution model construction unit is used to select an optimal Copula function based on the time series probability density model of wind power output and photovoltaic output, using the minimum square Euclidean distance as a correlation measure and fitting index to construct a wind-solar output time series joint probability distribution model; A sampling unit is used to perform Gibbs sampling based on the time series joint probability distribution model of wind and photovoltaic outputs to obtain a joint probability distribution sample group of wind power output and photovoltaic output, and substitute the sample group into the inverse function of the cumulative probability distribution of wind power output and photovoltaic output to determine a time series related sample group of wind power output and photovoltaic output; The system power supply capacity evaluation calculation unit is used to input the time-series related sample group of wind power output and photovoltaic output into the probabilistic power supply capacity evaluation calculation model, and use the adaptive step-size repeated power flow method to perform the system power supply capacity evaluation calculation; wherein, the probabilistic power supply capacity evaluation calculation model includes a solution model, equality constraints and inequality constraints; the solution model is constructed with the maximum system supply load as the objective function.

8. An electronic device, characterized in that: It comprises a memory and a processor, the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the sequential probabilistic power flow power supply capability assessment method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the sequential probabilistic power flow power supply capability assessment method as described in any one of claims 1 to 6.

Images