A consistent economic dispatch method based on non-coordinated step size in smart grid

By introducing a consistency algorithm of non-coordinated step size and row random matrix in directed unbalanced network, the problem of slow convergence of distributed economic scheduling algorithm in directed network is solved, and the rapid solution of economic scheduling problems is achieved. Improved quality.

CN114841830BActive Publication Date: 2024-11-26HEILONGJIANG ELECTRIC POWER SCIENCE RESEARCH INSTITUTE +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210419723.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-20
Publication Date
2024-11-26
Estimated Expiration
2042-04-20

AI Technical Summary

Technical Problem

The existing distributed economic scheduling algorithm converges slowly in directed imbalance networks, and the existing technology is difficult to effectively balance the imbalance of directed graphs, resulting in high communication costs and poor robustness.

Method used

A consistent economic scheduling method based on non-coordinated step size is proposed. Through mathematical modeling and minimizing the dual problem, attenuation step size and row random matrix are introduced, and a consistency algorithm is designed to balance directed imbalance networks to achieve economic Quick solution to scheduling problems.

Benefits of technology

This method can accelerate the optimal solution convergence speed of economic scheduling problems, reduce the complexity of step size adjustment, improve the robustness and efficiency of the system, and is suitable for economic scheduling problems with capacity constraints and supply and demand balance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114841830B_ABST
    Figure CN114841830B_ABST
Patent Text Reader

Abstract

The present invention claims protection for a consistent economic dispatch method based on non-coordinated step size in a smart grid, and belongs to the field of smart grid economic dispatch. Assuming the incremental cost of the generator as a consistent variable, a new non-coordinated step size distributed Lagrangian algorithm is proposed to speed up the convergence of the algorithm, and a row random matrix is ​​introduced to balance the imbalance of the directed network. Finally, the convergence of the algorithm is analyzed using the small gain theorem and the optimal solution to the economic dispatch problem is obtained. It is suitable for solving economic dispatch problems with capacity constraints and supply and demand balance constraints under directed unbalanced strong connected topology, and its power generation cost function is a convex function. While meeting the dispatching needs of the smart grid, the present invention can quickly obtain the optimal solution to the economic dispatch problem, and considering the directed unbalanced communication topology, it has a wider applicability and can ensure the safe and stable operation of the smart grid.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of smart grid economic dispatching, and relates to a distributed economic dispatching algorithm based on non-coordinated step lengths under directed unbalanced network topology. Background Art

[0002] Energy is vital to the development of human society, and energy shortage is an eternal problem. With the development trend of modern society and industry, global energy consumption is increasing, and people are generally concerned about how to save energy. Since electricity is the most consumed energy, it has attracted the attention of researchers. How to ensure the safe, economical and stable operation of the power system has become a research hotspot for researchers, thus giving birth to the classic problem in smart grids - the economic dispatch problem. The economic dispatch problem mainly solves the problem of supply and demand balance of generator sets in the power system and the problem of minimizing the total power generation cost. By reasonably dispatching the operating status of generator sets, improving the operating efficiency of the power system, and reducing the total power generation cost of the power system, it can effectively reduce energy loss and economic losses.

[0003] Traditionally, most economic dispatch problems are solved by adopting centralized solutions. Therefore, many centralized algorithms have emerged, such as Lagrange relaxation method, lambda iteration method, Newton method, etc. However, centralized solutions inevitably require a central processor to collect and process data information of the entire network, and its communication cost is very high. As a result, distributed algorithms came into being. In distributed algorithms, each generator set can communicate freely, and the failure of a single generator does not affect the operation of other generators, which has high robustness.

[0004] In recent years, the economic dispatch problem has been widely studied based on the consensus theory of multi-agent systems. In these algorithms, the incremental cost of the generator is usually selected as the consistency variable, and the optimal solution to the economic dispatch problem is obtained by communicating with other local generators. A common problem in solving economic dispatch algorithms on undirected networks is that their convergence speed is often not explicitly given, or they show a slower sublinear rate compared to centralized algorithms. For directed networks, due to the use of a decreasing step size method, its slower convergence speed is inevitable. Most of the existing work considers a constant step size, and the range of the constant step size obtained by theoretical analysis is usually affected by multiple constraints and needs to be adjusted manually. For directed unbalanced network communication topology, most of them use row random matrices and column random matrices to balance the imbalance of the directed graph, which requires each generator to know its out-degree information to adjust its out-degree weight and ensure that the weight matrix is ​​a column random matrix, but this does not meet practical applications.

[0005] Therefore, it is of great practical significance to design a distributed economic dispatch algorithm considering uncoordinated step sizes for directed unbalanced networks and to solve economic dispatch problems with capacity constraints and inequality constraints.

[0006] After searching, the application publication number CN111242513A, a method for distributed economic dispatch of power system based on consistency theory, includes selecting generator sets according to user demand power, selecting leader units from them, and building strong connection communication of each unit; mathematical modeling of economic dispatch of distributed power system, and giving optimized unit iteration rules and iteration initial values ​​based on mathematical modeling and strong connection communication topology; iterating according to the iteration rules and iteration initial values ​​until the sum of the mismatch power of each leader unit is less than the set threshold, and judging that the economic dispatch has reached the balance of supply and demand. During the iteration process, each unit not only feeds back its own mismatch power to the consistency variable of the unit, but also feeds back the mismatch power of all neighboring units communicating with it to the consistency variable of the unit, so that the economic dispatch of the power system can be more sensitive to the error perception of each unit, so as to achieve the balance of supply and demand of the power system more quickly. When solving the economic dispatch problem, this patent takes into account the directed balanced network topology and uses a leader-follower consensus algorithm. The leader-follower consensus algorithm is not essentially a fully distributed algorithm. Compared with the experimental results of the present invention, the convergence speed of the invention is slower. The present invention studies the economic dispatch problem based on a directed unbalanced network and introduces a non-coordinated step size to accelerate the solution of the economic dispatch problem. Therefore, the convergence speed is faster and it is more suitable for solving the economic dispatch problem. Summary of the invention

[0007] The present invention aims to solve the above problems of the prior art. A consistent economic dispatch method based on non-coordinated step length in a smart grid is proposed. The technical solution of the present invention is as follows:

[0008] A consistent economic dispatch method based on non-coordinated step size in a smart grid comprises the following steps:

[0009] Step 1: Mathematically model the economic dispatch problem in the power system. Based on the directed unbalanced strongly connected topological graph, the power generation cost function is assumed to be a convex function and smooth. The economic dispatch problem with supply and demand balance constraints and capacity constraints is considered. The model is as follows:

[0010]

[0011]

[0012]

[0013] Step 2: First transform the economic dispatch problem into the corresponding dual problem, and then transform it into the following minimization form based on the dual problem:

[0014] qi(y) represents the economic dispatch problem model;

[0015] Step 3: Considering the information interaction capability between generators, set the initial power value of the generator and the parameter a of the generator cost function i ,b i ,c i , and the corresponding generator weight r ij , where the generator weights satisfy:

[0016] in τ is a constant greater than 0, They represent the inner neighbor sets of generator i, i and j represent the i-th generator and the j-th generator respectively.

[0017] Step 4: According to the non-coordinated step size rule: η i (t) = ρα t , select appropriate parameter values ​​ρ, α, where ρ, α represent the parameters to be set, η i (t) represents the attenuation step length at time t;

[0018] Step 5: Update the corresponding consistency variables according to the consistency algorithm, iterate multiple times, and obtain the optimal solution to the dual problem;

[0019] Step 6: Convert the optimal solution of the dual problem into the optimal solution of the original problem, thereby obtaining the optimal solution x of the economic dispatch problem * And the corresponding convergence speed, output the optimal power value.

[0020] Furthermore, the step 1 mathematically models the economic dispatch problem in the power system, and the model is as follows:

[0021]

[0022]

[0023]

[0024] where f i (x i ) is the cost function of the generator, which is expressed in the following form: Among them, a i ,b i ,c i represents the cost coefficient, x iis the power generated by the ith generator, d i is the load demand of the generator, is the supply and demand balance constraint of the generator, and are the maximum and minimum values ​​of the generator, respectively, which are the capacity constraints of the generator.

[0025] Furthermore, the step 2 specifically includes: according to the modeling of step 1, since the cost function is strongly convex and smooth and satisfies the Slater condition, the Lagrangian function is introduced, and the corresponding dual problem of the economic dispatch problem is as follows:

[0026] Φ i (y) = ψ i (y)+yd i

[0027] in y is the Lagrange multiplier, d i Represents the load demand of the generator. In order to solve the optimal solution of the economic dispatch problem in a distributed manner and applicable to directed unbalanced networks, its dual problem is transformed into the following minimization form for solution:

[0028]

[0029] in

[0030] Furthermore, in step 4, the attenuation step length η i (t) satisfies: η i (t+1)<η i (t) and

[0031] Furthermore, step 5 sets a corresponding consistency algorithm according to the first four steps, and updates the corresponding consistency variables according to the consistency algorithm, iterating multiple times to obtain the optimal solution to the dual problem, which specifically includes:

[0032] Assume the incremental cost of the generator as the consistency variable, design the corresponding consistency algorithm, and update the corresponding consistency variable according to the consistency algorithm. The algorithm uses a row random matrix to balance the imbalance of the directed unbalanced network. After multiple iterations, the optimal solution to the dual problem can be obtained. The consistency algorithm is as follows:

[0033]

[0034]

[0035]

[0036] in, ηi (t) is the non-coordinated step length, [s i (t)] i Indicates i The i-th item of (t), Yes i (y) when y=y i The gradient at (t), the weight matrix R = {r ij} is row random, Indicates at time The consistency value of generator i.

[0037] Furthermore, the step 6: converting the optimal solution of the dual problem into the optimal solution of the original problem specifically includes: assuming that the directed topological graph is a strongly connected graph, the cost function of the generator is smooth and non-convex, the incoherent step size satisfies the rule of step 4, and the update rule of the consistency algorithm is as described in step 5. According to the small gain theorem, it can be deduced that the consistency variable y{t} converges to 1 n y * ,y * represents the optimal solution to the dual problem.

[0038] The optimal solution to the economic dispatch problem is obtained, and the Lagrangian function is introduced as follows:

[0039] L i (x i ,y i ) = f i (x i )+y i (x i -d i ).

[0040] Lagrangian function L i It is also strongly convex. Combined with the consistency algorithm update rule in step 5, since the consistency variable y{t} converges to 1 n y * ; It can be deduced that x(t) is moving at a linear rate O((γ / 2) t ) converges to x * , thus obtaining the optimal solution x of the economic dispatch problem * .

[0041] The advantages and beneficial effects of the present invention are as follows:

[0042] 1. The present invention considers the economic dispatch problem with capacity constraints and supply-demand balance constraints, and converts it into a minimization form of a dual problem for solution;

[0043] 2. The attenuation step length in claim 4 of the present invention changes with time and does not require manual adjustment, which is more flexible in the selection of the step length and can accelerate the optimal solution to the economic dispatch problem;

[0044] 3. The consistency algorithm in claim 5 of the present invention is derived from a row random matrix combined with claims 1-4. The use of a row random matrix can eliminate the imbalance of a directed unbalanced network. Compared with a double random matrix, it is easier to implement, and the optimal solution to the economic scheduling problem can be obtained through this consistency algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a communication topology diagram of a preferred embodiment provided by the present invention;

[0046] Figure 2 is a diagram of the generator coefficients of the present invention;

[0047] Figure 3 is a flow chart of the present invention;

[0048] Figure 4 is a graph showing changes in output power of the generator of the present invention;

[0049] Figure 5 It is the consistency variable change diagram of the present invention;

[0050] Figure 6 It is the supply and demand balance diagram of the present invention. DETAILED DESCRIPTION

[0051] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0052] The technical solution of the present invention to solve the above technical problems is:

[0053] Assume that the system contains N generators, 1, 2, ..., N. G = {V, ε} represents the communication topology between generators, where V = {v 1 ,v 2 ,...,v n} represents a vertex set, Directed edge set. Directed edge e ij =(i,j)∈ε means that node i can receive information from node j. represents the inner neighbor set of generator i, represents the outer neighbor aggregation of generator i, if Then the directed graph is an unbalanced directed graph, such as Figure 1 shown.

[0054] Step 1: Mathematically model the economic dispatch problem in the power system. Based on a directed strongly connected topological graph, assume that the power generation cost function is a convex function and is smooth, and consider the economic dispatch problem with supply and demand balance constraints and capacity constraints. The model is as follows:

[0055]

[0056]

[0057]

[0058] where f i (x i ) is the cost function of the generator, which is expressed in the following form: Among them, a i ,b i ,c i Represents the cost coefficient, and its related parameter values ​​are as follows Figure 2 As shown, x i is the power generated by the ith generator. i is the load demand of the generator, is the supply and demand balance constraint of the generator, and are the maximum and minimum values ​​of the generator, respectively, which are the capacity constraints of the generator.

[0059] Step 2: Based on the modeling in step 1, since the cost function is strongly convex and smooth and satisfies the Slater condition, the Lagrangian function is introduced, and the corresponding dual problem of the economic dispatch problem is as follows:

[0060] Φ i (y) = ψ i (y)+yd i

[0061] in In order to solve the optimal solution of the economic dispatch problem in a distributed manner and applicable to directed unbalanced networks, the dual problem is transformed into the following minimization form for solution:

[0062]

[0063] in

[0064] Step 3: Considering the information interaction capability between generators, set the initial power value x of the generator i (0) and the consistency variable y i (0), and the corresponding generator weight r ij The generator weights satisfy:

[0065] in

[0066] Step 4: According to the non-coordinated step size rule: η i (t) = ραt , select appropriate parameter values ​​ρ, α, where ρ, α are the parameters of the update rule, satisfying ρ>1,0<α<1, due to η i (t) is the attenuation step length, then η i (t+1)<η i (t) and

[0067] Step 5: Based on the above steps, let the incremental cost of the generator be the consistency variable, design the corresponding consistency algorithm, and update the corresponding consistency variable according to the consistency algorithm. The algorithm uses a row random matrix to balance the imbalance of the directed unbalanced network. After multiple iterations, the optimal solution to the dual problem can be obtained. The consistency algorithm is as follows:

[0068]

[0069]

[0070]

[0071] in, η i (t) is the non-coordinated step length.[s i (t)] i Indicates i The i-th item of (t), Yes i (y) when y=y i In addition, according to step 3, the weight matrix R = {r ij} is row random.

[0072] Step 6: Assume that the directed topology graph is a strongly connected graph, the cost function of the generator is smooth and non-convex, the incoherent step size satisfies the rule of step 4, and the update rule of the consistency algorithm is as described in step 5. According to the small gain theorem, it can be deduced that the consistency variable y{t} converges to 1 n y * .

[0073] Step 7: Based on the above results, the optimal solution of the dual problem is obtained, and the Lagrangian function is introduced as follows: L i (x i ,y i ) = f i (x i )+y i (x i -d i )

[0074] Lagrangian function L iIt is also strongly convex. Combined with the consistency algorithm update rule in step 5, since the consistency variable y{t} converges to 1 n y * It can be deduced that x(t) increases at a linear rate O((γ / 2) t ) converges to x * , thus obtaining the optimal solution x of the economic dispatch problem * Based on the above steps, assuming N=5, a simulation experiment is conducted on a directed unbalanced network with 5 generators. The flowchart of the present invention is as follows: Figure 3 The specific results are shown in Figure 4 , Figure 5 , Figure 6 As shown. Figure 4 , it can be seen that the power generation of each generator converges to the optimal solution of the economic dispatch problem at about 17s, where the optimal power is: Figure 5 shows the convergence curve of the consistency variable, the consistency variable y * =-36.89. Figure 6 Represents the supply and demand balance change curve between generators.

[0075] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0076] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A consistent economic dispatch method based on non-coordinated step size in smart grid. It is characterized in that The following steps are involved: Step 1: Mathematically model the economic dispatch problem in the power system. Based on the directed unbalanced strongly connected topological graph, the power generation cost function is assumed to be a convex function and smooth. The economic dispatch problem with supply and demand balance constraints and capacity constraints is considered. The model is as follows: Step 2: First transform the economic dispatch problem into the corresponding dual problem, and then transform it into the following minimization form based on the dual problem: qi(y) represents the economic dispatch problem model; Step 3: Considering the information interaction capability between generators, set the initial power value of the generator and the parameter a of the generator cost function i ,b i ,c i , and the corresponding generator weight r ij , where the generator weights satisfy: in τ is a constant greater than 0, They represent the inner neighbor sets of generator i, i and j represent the i-th generator and the j-th generator respectively; Step 4: According to the non-coordinated step size rule: η i (t) = ρα t , select appropriate parameter values ​​ρ, α, where ρ, α represent the parameters to be set, η i (t) represents the attenuation step length at time t; Step 5: Update the corresponding consistency variables according to the consistency algorithm, iterate multiple times, and obtain the optimal solution to the dual problem; Step 6: Convert the optimal solution of the dual problem into the optimal solution of the original problem, thereby obtaining the optimal solution x of the economic dispatch problem * And the corresponding convergence speed, output the optimal power value; Based on the directed unbalanced network topology, the economic dispatch problem in the power system is mathematically modeled according to step 1, and the model is as follows: where f i (x i ) is the cost function of the generator, which is expressed in the following form: Among them, a i ,b i ,c i represents the cost coefficient, x i is the power generated by the ith generator, d i is the load demand of the generator, is the supply and demand balance constraint of the generator, and are the maximum and minimum values ​​of the generator, i.e., the capacity constraints of the generator; The step 2 specifically includes: according to the modeling of step 1, since the cost function is strongly convex and smooth and satisfies the Slater condition, the Lagrangian function is introduced, and the corresponding dual problem of the economic dispatch problem is as follows: F i (y)=ψ i (y)+yd i in y is the Lagrange multiplier, d i Represents the load demand of the generator. In order to solve the optimal solution of the economic dispatch problem in a distributed manner and applicable to directed unbalanced networks, its dual problem is transformed into the following minimization form for solution: among themq i (y)=-Φ i (y)=f i ⊥ (y)-yd i ; In step 4, the attenuation step length η i (t) satisfies: η i (t+1)<η i (t) and The step 5 sets the corresponding consistency algorithm according to the first four steps, and updates the corresponding consistency variables according to the consistency algorithm, iterates multiple times, and obtains the optimal solution to the dual problem, which specifically includes: Assume the incremental cost of the generator as the consistency variable, design the corresponding consistency algorithm, and update the corresponding consistency variable according to the consistency algorithm. The algorithm uses a row random matrix to balance the imbalance of the directed unbalanced network. After multiple iterations, the optimal solution to the dual problem is obtained. The consistency algorithm is as follows: in, η i (t) is the non-coordinated step length, [s i (t)] i Indicates i The i-th item of (t), Yes i (y) when y=y i The gradient at (t), the weight matrix R = {r ij } is row random, Indicates at time The consistency value of generator i; Step 6: Convert the optimal solution of the dual problem into the optimal solution of the original problem, specifically including: assuming that the directed topological graph is a strongly connected graph, the cost function of the generator is smooth and non-convex, the incoherent step size satisfies the rule of step 4, the update rule of the consistency algorithm is as described in step 5, and according to the small gain theorem, the consistency variable y{t} converges to 1 n y * ,y * represents the optimal solution to the dual problem; The optimal solution to the economic dispatch problem is obtained, and the Lagrangian function is introduced as follows: L i (x i ,y i )=f i (x i )+y i (x i -d i ); Lagrangian function L i It is also strongly convex. Combined with the consistency algorithm update rule in step 5, since the consistency variable y{t} converges to 1 n y * ; It follows that x(t) moves at a linear rate O((γ / 2) t ) converges to x * , thus obtaining the optimal solution x of the economic dispatch problem * .