A new random state estimation method for distribution network

Abstract

The present invention relates to a novel random state estimation method for distribution network, comprising the following steps: 1) defining a state estimation loss function; 2) establishing a random state estimation model considering pseudo-measurement uncertainty; 3) using Lagrange multiplier method to solve the random state estimation model. Compared with the prior art, the present invention has the advantages of high calculation accuracy and high speed.

Description

Technical Field

[0001] The present invention relates to the technical field of novel distribution network state estimation, and in particular to a distribution network state estimation method taking into account new energy and load measurement uncertainty. Background Art

[0002] As the scale of new energy access to the distribution network continues to increase, the distribution network faces increasing uncertainty due to its natural randomness, especially for new energy without measurement configuration. In order to ensure the observability of state estimation, its predicted value is often used as pseudo-measurement, which makes the measurement error have a large uncertainty. How to improve the accuracy of state estimation calculation under the influence of uncertainty factors has become an urgent problem to be solved.

[0003] Uncertainty quantification methods have been widely used in random power flow and economic dispatch of power systems, mainly including the following:

[0004] (1) Monte Carlo simulation and its improved algorithm. The idea is to use the probability distribution function of the uncertainty variable for sampling, and then perform deterministic calculations on each group of samples to obtain the value of the output state quantity corresponding to each group of samples, and then obtain the statistical characteristics of the output state quantity to be analyzed based on these values. It mainly includes the important sampling method, Latin hypercube sampling method, quasi-Monte Carlo method and hybrid methods of these methods. The main disadvantage of this type of method is the large amount of calculation. In order to obtain higher accuracy, it is often necessary to sample a large number of sample points, which is difficult to meet engineering applications. Its results are usually used as a comparison of the accuracy of other uncertainty quantification methods.

[0005] (2) Approximate method, the idea is to use the digital characteristics (low-order moments) of the input random variables to approximate the statistical characteristics (mean and variance) of the system output state quantity. It mainly includes point estimation method, first-order second-order moment method and state transformation method, etc. Its disadvantage is that the accuracy is not high, especially the high-order moments of the output state quantity, so it is impossible to accurately obtain the overall probability distribution of the output state quantity.

[0006] (3) Analytical method: This method uses the independent distribution characteristics between random variables to obtain the probability distribution of the output state quantity through convolution calculation. It mainly includes fast Fourier transform, semi-invariant method and sequence operation theory. Among them, fast Fourier transform and sequence operation theory are difficult to apply due to the large amount of convolution calculation. The idea of ​​this type of method is to calculate the semi-invariant of each order of the output state quantity based on a certain deterministic power flow calculation result, so as to obtain the probability distribution of the output state quantity. Its disadvantage is that when the uncertainty variable fluctuates greatly, its calculation accuracy is not high.

[0007] (4) Interval optimization method: This method is an optimization method that is performed when the probability distribution of random variables is unknown or inaccurate, and only the interval information of random variables is known. The result obtained by this method is an interval result, and how to provide dispatchers or operators with a clear optimal solution is still a problem to be solved.

[0008] (5) The generalized polynomial chaos method is a semi-analytical method that has emerged in recent years. Its main idea is to establish a polynomial function relationship between the uncertainty variable and the output state quantity, and then calculate the coefficients of the polynomial by random point matching method (such as sparse grid method), so as to obtain the random characteristics of the output state quantity. The unity of calculation speed and calculation accuracy can be achieved by controlling the degree of the polynomial. In the early stage of research, this method was limited to analyzing the impact of a single uncertainty factor on state estimation. So far, the application of this method in the engineering field has been expanded to the analysis and calculation of multiple uncertain variables, and the calculation accuracy and speed have been greatly improved. However, this method fails to answer a question, that is, how reliable the calculation results are.

[0009] (6) Robust optimization method: The key to this method is to establish a corresponding robust equivalent model. Due to the mathematical characteristics of the robust optimization model, it is often necessary to approximate the nonlinear model as a quadratic or linear model. However, engineering problems often have strong nonlinearity. In addition, the method is relatively conservative in its solution, which limits its engineering application.

[0010] (7) Conditional risk method: Its main idea is to establish a loss function and transform the loss function into uncertainty risk through the conditional risk model. The calculation of risk involves the integration of the loss function, which is then approximated by the Monte Carlo sampling method. As an effective tool in finance to characterize the size of the risk brought by uncertainty factors, the conditional risk method has also been widely used in recent years to quantify the risk caused by randomness in the field of power engineering. It can provide the decision variable value and confidence under the influence of random factors, and its physical meaning is relatively clear. Summary of the invention

[0011] The purpose of the present invention is to provide a novel random state estimation method for distribution network in order to overcome the defects of the above-mentioned prior art.

[0012] The purpose of the present invention can be achieved by the following technical solutions:

[0013] A novel random state estimation method for distribution network comprises the following steps:

[0014] Step 1) define the state estimation loss function;

[0015] Step 2) establishing a random state estimation model considering pseudo-measurement uncertainty;

[0016] Step 3) Use the Lagrange multiplier method to solve the random state estimation model.

[0017] Furthermore, the state estimation loss function is:

[0018]

[0019] Among them, m ξ is the total number of pseudo measurements in the distribution network, m is the total number of measurements in the distribution network, ξ i is the i-th sample of random variables ξ such as wind power and photovoltaic power generation connected to the distribution network for Monte Carlo simulation, w i For distribution network measurement i The weight, h i (x) is z i The calculated value of , x is the state variable of the distribution network.

[0020] Furthermore, the random state estimation model established in step 2) is:

[0021]

[0022] stf(x,ξ j )-y0-y j ≤0

[0023] y j ≥0

[0024] j=1,2,…,n

[0025] y0∈R;

[0026] Among them, ξ j For the jth group m ξ is the random variable sampling of pseudo-measurement of the distribution network, α is the confidence level, n is the number of samples sampled by Monte Carlo simulation for random variables ξ connected to the distribution network, y i (i=0,1,2,…,n) is an auxiliary variable, and R represents a real number.

[0027] Furthermore, the step 3) specifically includes the following steps:

[0028] Step 31) Construct the Lagrangian function and convert the random state estimation model into an unconstrained optimization problem:

[0029]

[0030] Among them, λ j ,y j ,μ j is a Lagrange multiplier, and λ j ,y j ≥0(j=1,2,…,n),gj (x,y)=f(x,ξ j )-y0-y j .

[0031] Step 32) Calculate the first-order partial derivatives of the state variables of the Lagrangian function to obtain the extreme value conditions:

[0032]

[0033] Among them, W k is the weight matrix,

[0034] By using Newton's method to solve the above simultaneous equations, the state of the distribution network can be obtained.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] 1. High calculation accuracy: Under the same convergence conditions, the calculation accuracy of the present invention is exactly the same as that of the Monte Carlo method.

[0037] 2. Fast calculation speed: To achieve the same calculation accuracy, the number of samples required by the present invention is only 1 / 25 of that of the Monte Carlo method, and the calculation time is one quarter of that of the Monte Carlo method, which greatly reduces the calculation time.

[0038] 3. Good application prospects: The new power system characterized by large-scale access to the grid of distributed power sources, electric vehicles, etc. is increasingly affected by strong uncertain factors such as intermittent new energy and diversified user needs. How to achieve comprehensive and accurate control of the operating status of the distribution network with uncertain access including renewable energy is crucial. The new distribution network random state estimation method involved in the present invention can provide good technical support for it, and therefore has good application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The figure is a flow chart of the method of the present invention.

[0040] Figure 2 This is a comparison diagram of the voltage amplitude estimation results of the method of the present invention.

[0041] Figure 3 This is a comparison diagram of the voltage phase angle estimation results of the method of the present invention. DETAILED DESCRIPTION

[0042] The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

[0043] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0044] like Figure 1 As shown, the present invention provides a random state estimation method considering pseudo-measurement uncertainties of new energy sources and loads in a distribution network, comprising the following steps performed in sequence:

[0045] Step 1) Define the state estimation loss function:

[0046]

[0047] Among them, m ξ is the total number of pseudo measurements in the distribution network, m is the total number of measurements in the distribution network, ξ i is the i-th sample of random variables ξ such as wind power and photovoltaic power generation connected to the distribution network for Monte Carlo simulation, w i For distribution network measurement i The weight, h i (x) is z i The calculated value of , x is the state variable of the distribution network.

[0048] Step 2) Establish a random state estimation model:

[0049]

[0050] stf(x,ξ j )-y0-y j ≤0

[0051] y j ≥0

[0052] j=1,2,…,n

[0053] y0∈R

[0054] Among them, ξ j For the jth group m ξ dimensional pseudo-measurement random quantity sampling, α is the confidence level, n is the number of samples sampled for the random variable ξ by Monte Carlo simulation, y i(i=0,1,2,…,n) is an auxiliary variable, and R represents a real number.

[0055] Step 3) The Lagrange multiplier method is used to solve the random state estimation model. The solution steps are as follows:

[0056] 31) Construct the Lagrangian function and transform the random state estimation model into an unconstrained optimization problem:

[0057]

[0058] Among them, λ j ,y j ,μ j is a Lagrange multiplier, and λ j ,y j ≥0(j=1,2,…,n),g j (x,y)=f(x,ξ j )-y0-y j .

[0059] 32) Calculate the first-order partial derivatives of the state variables of the Lagrangian function and obtain the extreme value conditions:

[0060]

[0061] Among them, W k is the weight matrix, By using Newton's method to solve the above set of simultaneous equations, the state estimation result of the distribution network can be obtained.

[0062] The effectiveness of the algorithm was tested using an IEEE 9-node system. The system has a total of 22 content measurements, among which the injected active power of nodes 3, 5, and 6 is set as pseudo-measurement. Node 3 is a wind power generation node. It is assumed that the wind speed obeys the Weibull distribution with a scale parameter of 8 and a shape parameter of 3. The cut-in, rated, and cut-out wind speeds are 2.5, 15, and 25 m / s, respectively, and their means are 0.7 times the actual values. Nodes 5 and 6 are load pseudo-measurement points that obey a normal distribution with means of 0.7 and 1.3 times the actual values, respectively, and a standard deviation of 0.2 times the actual value. The estimation accuracy of the state variables of the present invention (CVaR) and the Monte Carlo method (MCM) are compared. The number of Latin hypercube sampling sample points used by CVaR is 4000, the confidence level α is 0.99, and the Monte Carlo method sampling sample is 10 5 .

[0063] From the attached Figure 2 , 3It can be seen that the calculation accuracy of CVaR and MCM is basically the same, which verifies that CVaR can effectively quantify the impact of pseudo-measurement randomness. In terms of calculation time, since CVaR has much fewer sample points than MCM, its calculation time is only 4.23s, while MCM takes 17.21s. This shows that the present invention has obvious advantages over traditional methods in calculation accuracy and calculation time.

[0064] It is not difficult to see from the above embodiments that the new distribution network random state estimation method proposed in the present invention has the advantages of high calculation accuracy and high speed. The specific embodiments described above are only for illustrating the implementation effect of the present invention and are not intended to limit the present invention. Any non-substantial modification, conversion and improvement made within the basic idea and framework of the method proposed in the present invention shall be included in the protection scope of the present invention.

[0065] In the description of the present invention, it is necessary to understand that the terms "middle", "length", "up", "down", "front", "back", "vertical", "horizontal", "inside", "outside", "radial", "circumferential" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0066] In the present invention, unless otherwise clearly specified and limited, the first feature "on" the second feature may be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. "Multiple" means at least two, such as two, three, etc., unless otherwise clearly and specifically limited.

[0067] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection, an electrical connection, or communication with each other; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0068] The above is only for explaining the implementation mode of the present invention and is not intended to limit the present invention. For those skilled in the art, any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention without creative work should be included in the protection scope of the present invention.

Claims

1. A novel random state estimation method for distribution network, characterized in that: The following steps are involved: Step 1) define the state estimation loss function; Step 2) establishing a random state estimation model considering pseudo-measurement uncertainty; Step 3) using Lagrange multiplier method to solve the random state estimation model; The state estimation loss function is: Among them, m ξ is the total number of pseudo measurements in the distribution network, m is the total number of measurements in the distribution network, ξ i is the i-th sample of random variables ξ such as wind power and photovoltaic power generation connected to the distribution network for Monte Carlo simulation, w i For distribution network measurement i The weight, h i (x) is z i The calculated value of , x is the state variable of the distribution network; The random state estimation model established in step 2) is: s.t.f(x,ξ j )-y0-y j ≤0 y j ≥0 j=1,2,…,n y0∈R; Among them, ξ j For the jth group m ξ is the random variable sampling of pseudo-measurement of the distribution network, α is the confidence level, n is the number of samples sampled by Monte Carlo simulation for random variables ξ connected to the distribution network, y i is an auxiliary variable, i=0,1,2,…,n, and R represents a real number.

2. A novel random state estimation method for distribution network according to claim 1, characterized in that: The step 3) specifically comprises the following steps: Step 31) Construct the Lagrangian function and convert the random state estimation model into an unconstrained optimization problem: Among them, λ j ,y j ,μ j is a Lagrange multiplier, and λ j ,y j ≥0,j=1,2,…,n,g j (x,y)=f(x,ξ j )-y0-y j ; Step 32) Calculate the first-order partial derivatives of the state variables of the Lagrangian function to obtain the extreme value conditions: Among them, W k is the weight matrix, By using Newton's method to solve the above simultaneous equations, the state of the distribution network can be obtained.

Images