Power distribution network frame optimization method based on quantum annealing algorithm

Abstract

The invention relates to a power distribution network frame optimization method based on a quantum annealing algorithm. The method comprises the following steps: constructing an objective function of power distribution network frame optimization based on actual power distribution network data and setting constraints; obtaining a quantum Hamiltonian function based on the objective function and the constraint; based on the quantum Hamiltonian function, obtaining a secondary unconstrained binary optimization objective function of power distribution network frame optimization; based on the quantum Hamiltonian function or the quadratic unconstrained binary optimization objective function, solving a power distribution network frame optimization optimal solution; and generating a power distribution network frame optimization strategy based on the optimal solution. Compared with the prior art, the method has the advantages of higher calculation efficiency and precision and the like.

Description

Technical Field

[0001] The present invention relates to the field of distribution network optimization, and in particular to a distribution network framework optimization method based on a quantum annealing algorithm. Background Art

[0002] The distribution network is a network structure in the power system that is responsible for transmitting electric energy from the substation to the user end. The lines in the grid structure directly connect the power station and the end user. The design and structure of the distribution network not only affect the efficiency and stability of power transmission, but also the power supply quality of the power grid and the user experience. The distribution network is usually composed of distribution lines, substation equipment, branch nodes, protection devices, etc., covering cities, suburbs and even remote areas. Therefore, it is of great significance to ensure the safe, stable and economical operation of the distribution network.

[0003] At present, the methods for solving the distribution network optimization problem can be mainly divided into the following two types:

[0004] (1) Mathematical optimization method

[0005] Classic mathematical optimization algorithms for solving distribution network optimization include linear programming (LP), integer programming (IP), dynamic programming (DP), etc. These methods transform the distribution network optimization problem into a solvable optimization problem by establishing a mathematical model to obtain the optimal grid structure solution. The linear programming method has a fast solution speed, but for large-scale complex optimization problems, it may require a lot of computing resources and has limited applicability; the integer programming method is suitable for discrete variables and cannot guarantee the global optimal solution; the dynamic programming method can find the global optimum, but it is limited by the state space dimension. For large-scale complex scenarios, the computational complexity increases sharply and is difficult to apply.

[0006] (2) Heuristic Optimization Method

[0007] At present, common heuristic optimization algorithms include simulated annealing algorithm, genetic algorithm, ant colony algorithm, particle swarm algorithm, etc. The simulated annealing algorithm is simple to implement and has a strong ability to jump out of the local optimum, but the convergence speed is slow; the genetic algorithm has strong adaptability, but requires multiple generations of evolution, high computational complexity, and long time consumption; the ant colony algorithm has good global search performance, but relies on appropriate parameter design and heuristic function design; the particle swarm algorithm is simple and easy to implement, with high computational efficiency, but it is easy to fall into the local optimum and has a limited scope of application.

[0008] In summary, both mathematical optimization methods and heuristic optimization algorithms are very mature and can show certain advantages in dealing with optimization problems. However, they still have defects such as slow convergence and easy falling into local optimality in the actual solution process.

[0009] Therefore, providing a method that can speed up the solution process and improve the accuracy of finding the global optimal solution is a problem that needs to be solved. Summary of the invention

[0010] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a distribution network optimization method based on quantum annealing algorithm.

[0011] The purpose of the present invention can be achieved by the following technical solutions:

[0012] According to a first aspect of the present invention, a distribution network optimization method based on a quantum annealing algorithm is provided, the method comprising:

[0013] Construct the objective function of distribution network optimization based on actual distribution network data and set constraints;

[0014] Obtaining a quantum Hamiltonian function based on the objective function and the constraints;

[0015] Based on the quantum Hamiltonian function, a quadratic unconstrained binary optimization objective function for distribution network optimization is obtained;

[0016] Solving the optimal solution for distribution network optimization based on the quantum Hamiltonian function or the quadratic unconstrained binary optimization objective function;

[0017] Generate distribution network optimization strategy based on the optimal solution.

[0018] As a preferred technical solution, the objective function is determined by the line investment and depreciation cost and the line power loss cost, and its expression is:

[0019]

[0020] Where f(x) represents the objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C0 represents the unit price of the line; τ max Indicates the time of maximum load of the power distribution system; ΔP irepresents the power loss of branch i; ΔI represents the total value of all branch current exceeding the limit; C1 is the branch current exceeding the limit penalty coefficient; ΔU is the total value of all load point voltage exceeding the limit; C2 is the load point voltage exceeding the limit penalty coefficient.

[0021] As a preferred technical solution, the constraints include branch current constraints, load point current constraints, grid radiation constraints and grid connectivity constraints, wherein the branch current constraint is that the current flowing through the branch does not exceed the maximum current allowed to pass through the branch; the load point current constraint is greater than the load point voltage lower limit and less than the load point voltage lower limit; the grid radiation constraint is that the path from the power source point to any load point is unique and has no loops; the grid connectivity constraint is that the power source point is connected to any load point.

[0022] As a preferred technical solution, the method for obtaining the quantum Hamiltonian function is to rewrite the objective function according to the grid radiation constraint and the grid connectivity constraint to obtain the quantum Hamiltonian function, and the expression of the quantum Hamiltonian function is:

[0023]

[0024] Where F(x) represents the quantum Hamiltonian function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C0 represents the unit price of the line; τ max Indicates the time of maximum load of the power distribution system; ΔP i represents the power loss of branch i; ΔI represents the total value of all branch current exceeding the limit; C1 is the branch current exceeding the limit penalty coefficient; ΔU is the total value of all load point voltage exceeding the limit; C2 is the load point voltage exceeding the limit penalty coefficient; C3 represents the penalty coefficient of the grid radiation constraint; C4 represents the penalty coefficient of the grid connectivity constraint; C5 represents the no-loop penalty coefficient; L3 represents the penalty function of the grid radiation constraint; L4 represents the penalty function of the grid connectivity constraint; L5 represents the no-loop penalty function; Γ(t) represents the magnetic field strength; represents the potential energy term of the Hamiltonian function.

[0025] As a preferred technical solution, the method for obtaining the secondary unconstrained binary optimization objective function of the distribution network grid optimization is to extract the line power loss cost calculation item based on the quantum Hamiltonian function, that is, The branch current constraint calculation term, i.e., C1|ΔI|, and the load point current constraint calculation term, i.e., C2|ΔU, are used as power flow constraints. The line investment and depreciation cost calculation terms based on the quantum Hamiltonian function, i.e., The grid radiation constraint calculation item, namely C3L3+C4L4, and the grid connectivity constraint calculation item, namely C5L5, generate a quadratic unconstrained binary optimization objective function, the expression of which is:

[0026]

[0027] Where F'(x) represents the quadratic unconstrained binary optimization objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C3 represents the penalty coefficient of the grid radiation constraint; C4 represents the penalty coefficient of the grid connectivity constraint; C5 represents the no-loop penalty coefficient; L3 represents the penalty function of the grid radiation constraint; L4 represents the penalty function of the grid connectivity constraint; L5 represents the no-loop penalty function.

[0028] As a preferred technical solution, the method for solving the optimal solution for distribution network optimization based on the quantum Hamiltonian function is:

[0029] Initialize the magnetic field strength in the quantum Hamiltonian function, the initial temperature of the algorithm, the number of inner loop iterations, and the outer loop cooling coefficient;

[0030] Calculate the initial solution and calculate the objective function value corresponding to the initial solution, take the objective function value corresponding to the initial solution as the initial global optimal solution, and calculate the initial Hamiltonian function value and initial potential energy term value corresponding to the initial global optimal solution;

[0031] Generate a random disturbance, calculate the disturbance solution under the disturbance and calculate the objective function value corresponding to the disturbance solution, and calculate the disturbance Hamiltonian function value and the disturbance potential energy term value based on the objective function value corresponding to the disturbance solution;

[0032] Calculate the Hamiltonian difference between the initial Hamiltonian function value and the perturbed Hamiltonian function value, and calculate the potential energy difference between the initial potential energy term value and the perturbed potential energy term value;

[0033] If the Hamiltonian difference and the potential energy difference are both greater than or equal to the preset value, the disturbance is not accepted and a random disturbance is regenerated;

[0034] If the Hamiltonian difference and the potential energy difference are both less than the preset values, the disturbance is accepted as the local optimal solution and the corresponding objective function value, and it is determined whether the termination condition is met. If not, the iterative calculation is continued;

[0035] If it is satisfied, the current local optimal solution at the time of algorithm termination is taken as the global optimal solution, and the global optimal solution and the corresponding objective function value are output. The global optimal solution is the optimal solution for distribution network optimization.

[0036] As a preferred technical solution, the preset value is 0.

[0037] As a preferred technical solution, the termination conditions include: the number of inner loop iterations reaches a maximum number of iterations and the magnetic field strength in the outer loop is less than a preset magnetic field strength.

[0038] As a preferred technical solution, the method for solving the optimal solution for distribution network optimization based on the quadratic unconstrained binary optimization objective function is:

[0039] Calculate the initial solution and use power flow to calculate the line power loss cost value corresponding to the initial solution;

[0040] Generate a quantum Hamiltonian function corresponding to an initial solution, and generate a corresponding quadratic unconstrained binary optimization objective function based on the quantum Hamiltonian function corresponding to the initial solution;

[0041] Embed the corresponding quadratic unconstrained binary optimization objective function into a D-Wave quantum computer to solve the current optimal solution;

[0042] Calculate all power flow constraints except the line power loss cost corresponding to the current optimal solution. If all constraints are satisfied, take the current optimal solution as the optimal solution for distribution network optimization. Calculate the line investment and depreciation cost based on the optimal solution for distribution network optimization. Calculate the total cost value in combination with the line power loss cost value of the optimal solution for distribution network optimization. Output the optimal solution for distribution network optimization and the corresponding total cost value.

[0043] If not, regenerate the solution and repeat the above calculation until the current optimal solution satisfies all power flow constraints.

[0044] As a preferred technical solution, the method for obtaining the initial solution is to calculate a solution that satisfies both the grid radiation constraint and the grid connectivity constraint, and the solution is the initial solution.

[0045] Compared with the prior art, the present invention has the following advantages:

[0046] 1) The present invention adopts quantum annealing algorithm to solve the distribution network optimization problem. Compared with the traditional optimization algorithm, which is easy to fall into the local optimal solution, the global optimal solution is obtained through quantum tunneling effect, which effectively avoids falling into the local optimal solution, speeds up the solution process and improves the optimization accuracy.

[0047] 2) Based on the quantum Hamiltonian function, the present invention constructs a quadratic unconstrained binary optimization objective function suitable for the D-Wave dedicated quantum computer, which retains the quantum effect in the calculation process. Compared with traditional calculation methods, its calculation accuracy is higher and fewer iterations are required, which improves the calculation efficiency and the accuracy of obtaining the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of the distribution network framework optimization method based on quantum annealing algorithm of the present invention;

[0049] Figure 2 It is a flow chart of a method flow chart for solving the optimal solution of distribution network grid optimization based on a quadratic unconstrained binary optimization objective function of the present invention;

[0050] Figure 3 The present invention provides a flow chart for solving the optimal solution for distribution network optimization based on the quantum Hamiltonian function. DETAILED DESCRIPTION

[0051] 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 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 should fall within the scope of protection of the present invention.

[0052] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent.

[0053] The reference to "embodiment" in this embodiment means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the embodiment. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiment described in this embodiment can be combined with other embodiments without conflict.

[0054] Distribution network optimization mainly focuses on how to reasonably configure and optimize the structure and resource distribution of the distribution network to improve the operating efficiency of the power grid, reduce energy consumption and operating costs, and ensure the safe, stable and economical operation of the distribution system. In the optimization of the distribution network, the optimization algorithm can be used to determine the optimal line selection to minimize the line investment, depreciation cost and line power loss cost. However, traditional intelligent algorithms often face the problems of slow convergence and easy to fall into local optimal solutions when dealing with optimization problems. For this reason, this embodiment proposes a distribution network optimization method based on a quantum annealing algorithm.

[0055] The quantum annealing algorithm is an optimization algorithm based on the principles of quantum mechanics. It can effectively solve complex combinatorial optimization problems. The quantum annealing algorithm is applied to the optimization of the distribution network framework. The quantum tunneling effect is used to improve the convergence speed and jump out of the local optimum, so that the global optimal solution can be found more effectively. This embodiment provides a new solution method for the optimization of the distribution network framework, which is of great significance for achieving the economic, environmental and sustainable development of the distribution network.

[0056] This embodiment provides a distribution network optimization method based on a quantum annealing algorithm. Specifically, the process of the method is as follows: Figure 1 As shown, the detailed steps include:

[0057] S1. Construct the objective function of distribution network optimization based on actual distribution network data and set constraints.

[0058] The optimization of distribution network is to supply power to several load points from the substation. The objective function is to establish a mathematical model of distribution network optimization by taking into account branch current constraints, load point voltage constraints, grid radiation constraints and grid connectivity constraints. The branch current constraint is that the current flowing through the branch does not exceed the maximum current allowed to pass through the branch, that is, I i max , I i represents the current in branch i, I max Indicates the maximum current allowed to pass through the branch current; the load point current is constrained to be greater than the load point voltage lower limit and less than the load point voltage lower limit, that is, U min i max , U min Indicates the lower limit of load point voltage, U i Voltage of branch i, U max It represents the upper limit of the voltage at the load point; the grid radiation constraint is that the path from the power source point to any load point is unique and has no loops; the grid connectivity constraint is that the power source point is connected to any load point.

[0059] ​​​For a line in a distribution network structure, there are only two situations: construction and non-construction. Construction is regarded as adding the line to the network, while non-construction is regarded as not adding the line to the network. Therefore, it can be encoded in binary coding, that is, when the i-th line joins the network x i is 1; when the i-th line is not added to the network x i Then it is 0. In detail, when optimizing the distribution network, the objective function includes line investment and depreciation cost and line power loss cost, and its expression is:

[0060]

[0061] f(x) represents the objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C0 represents the unit price of the line; τ max Indicates the time of maximum load of the power distribution system; ΔP i represents the power loss of branch i; ΔI represents the total value of all branch current exceeding the limit; C1 is the branch current exceeding the limit penalty coefficient; ΔU is the total value of all load point voltage exceeding the limit; C2 is the load point voltage exceeding the limit penalty coefficient.

[0062] S2. Obtaining a quantum Hamiltonian function based on the objective function and constraints.

[0063] The model of the quantum annealing algorithm consists of two parts, one is quantum potential energy, and the other is quantum kinetic energy; quantum potential energy maps the optimization problem to the quantum system, and maps the optimized objective function to a potential field; quantum kinetic energy controls the penetration field of quantum fluctuations. These two parts can constitute the evaluation function under the quantum system, that is, the quantum Hamiltonian function. The quantum annealing process is usually simulated using a random process. This embodiment uses the path integral Monte Carlo method to simulate the random process of the quantum annealing algorithm, and its expression is:

[0064] H(t)=H pot (t)+H kin (t),

[0065] H(t) is the quantum Hamiltonian function, H pot (t) represents the quantum potential energy term, H kin (t) represents the quantum kinetic energy term.

[0066] S21. In this embodiment, the method for obtaining the quantum potential energy term is to rewrite the objective function according to the grid radiation constraint and the grid connectivity constraint to obtain the quantum potential energy term. The detailed steps are as follows:

[0067] A1. Construct a tree structure to meet the grid radiation constraints:

[0068] For a distribution network, connectivity and no loops need to be satisfied. To satisfy the interconnection and no loops, the tree structure must be satisfied first. For a distribution network grid structure diagram including n nodes, it must only include n-1 edges; therefore, the tree structure controls the number of lines selected in the grid. If a distribution network grid structure includes n nodes and m branches, then the form of its penalty function is:

[0069] C3(x1+x2+...+x m -(n-1)) 2 ,

[0070] C3 represents the penalty coefficient of grid radiation constraint, x m represents branch m, and n represents the nth node.

[0071] A2. Constructing connectivity constraints of tree structure:

[0072] For a distribution network structure, each node in the grid must have at least one branch connected to other nodes. For example, for a node connected by branch x1 and branch x2, the connectivity constraint at the node can be expressed as: x1+x2≥1, which is converted into a penalty function expression:

[0073] C4(1-x1-x2+x1*x2) 2 ,

[0074] C4 represents the penalty coefficient of grid connectivity constraint.

[0075] According to step A2, all nodes in the distribution network grid are constrained as above, and the connectivity constraints of all nodes are summed up to obtain the connectivity constraints of the distribution network grid.

[0076] A3. Construct acyclic constraints to achieve grid radiation constraints on the basis of satisfying connectivity:

[0077] When the grid only meets the connectivity, loop problems cannot be avoided during the connection between nodes and branches, and the radiation constraints of the grid structure cannot be met. Therefore, it is necessary to further construct loop-free constraints to ensure the grid radiation constraints.

[0078] For example, branches x1, x2, x3, and x4 form a loop. In order to prevent all branches from being selected during the algorithm, it is necessary to establish an acyclic constraint: x1+x2+x3+x4≤3. In the process of constructing the constraint, it is necessary to introduce a slack variable s to convert it into an equality constraint: x1+x2+x3+x4+s=3. Since the decision variable x of branch construction takes the value of 0 or 1, s can be represented by at most two binary bits, which can be written as: x1+x2+x3+x4+x5+2x6=3. Finally, it can be converted into the following penalty function:

[0079] C5(x1+x2+x3+x4+x5+2x6-3) 2 ,

[0080] C5 represents the loop-free penalty coefficient.

[0081] According to step A3, all key nodes are constructed with loop-free constraints, and the loop-free constraints of all key nodes are summed up to obtain the loop-free constraints of the distribution network grid, wherein the key nodes include nodes or power supply points that can form multiple loops.

[0082] According to the penalty function form of the grid radiation constraint and grid connectivity constraint of the distribution network grid structure in steps A1-A3, and writing it into the objective function, the quantum potential energy term /

[0083] S22. Construct the quantum kinetic energy term of the quantum Hamiltonian function based on the line information in the distribution network grid, and its expression is:

[0084]

[0085] Γ(t) represents the magnetic field strength.

[0086] S23. Combine the quantum potential energy term and the quantum kinetic energy term to obtain the quantum Hamiltonian function.

[0087] S3. Obtain the quadratic unconstrained binary optimization objective function based on the quantum Hamiltonian function.

[0088] The D-Wave dedicated quantum computer is a computer that uses the quantum annealing algorithm to solve combinatorial optimization problems. Its basic unit is the quantum bit. The quantum annealing algorithm calculated by the D-Wave quantum computer truly has quantum effects and focuses more on the objective function itself, without considering some parameters in the quantum annealing algorithm. Compared with classical computers, it can find the global optimum more quickly in a complex solution space. However, the input of the D-Wave quantum computer is a quadratic unconstrained binary optimization objective function, so the quantum Hamiltonian function needs to be converted into a QUBO model.

[0089] S31. The quantum Hamiltonian function already satisfies the objective function form of quadratic unconstrained binary optimization. However, the D-Wave quantum computer can only process fixed functions. It cannot calculate the uncertain line power loss cost, branch current constraints, and load point current constraints in the grid structure to be optimized. Therefore, it is necessary to remove the line power loss cost calculation item in the quantum Hamiltonian function. The branch current constraint calculation item, namely C1|ΔI|, and the load point current constraint calculation item, namely C2|ΔU|, are used as power flow constraints to solve the power flow calculation.

[0090] S32, extract the line investment and depreciation cost calculation items in the quantum Hamiltonian function, that is, The grid radiation constraint calculation item, i.e., C3L3+C4L4, and the grid connectivity constraint calculation item, i.e., C5L5, generate a quadratic unconstrained binary optimization objective function. Specifically, the expression of the quadratic unconstrained binary optimization objective function is:

[0091]

[0092] F'(x) represents the quadratic unconstrained binary optimization objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C3 represents the penalty coefficient of the grid radiation constraint; C4 represents the penalty coefficient of the grid connectivity constraint; C5 represents the no-loop penalty coefficient; L3 represents the penalty function of the grid radiation constraint; L4 represents the penalty function of the grid connectivity constraint; L5 represents the no-loop penalty function.

[0093] S4. Solve the optimal solution for distribution network optimization based on the quadratic unconstrained binary optimization objective function. The process is as follows Figure 2 As shown, the detailed steps include:

[0094] S41, calculating an initial solution, and performing power flow calculation on the line power loss cost value corresponding to the initial solution.

[0095] The method of the initial solution is to calculate a solution that satisfies both the grid radiation constraint and the grid connectivity constraint, and the solution is the initial solution.

[0096] S42, generating a quantum Hamiltonian function corresponding to the initial solution, and generating a corresponding quadratic unconstrained binary optimization objective function based on the quantum Hamiltonian function corresponding to the initial solution.

[0097] S43. Embed the corresponding quadratic unconstrained binary optimization objective function into the D-Wave quantum computer to solve the current optimal solution.

[0098] S44. Calculate all power flow constraints except the line power loss cost corresponding to the current optimal solution. If all constraints are satisfied, use the current optimal solution as the optimal solution for distribution network grid optimization. Calculate the line investment and depreciation cost based on the optimal solution for distribution network grid optimization. Calculate the total cost value in combination with the line power loss cost value of the optimal solution for distribution network grid optimization. Output the optimal solution for distribution network grid optimization and the corresponding total cost value.

[0099] S45. If not, regenerate the solution and repeat the calculation until the current optimal solution satisfies all power flow constraints.

[0100] S5. Generate a distribution network optimization strategy based on the optimal solution.

[0101] This embodiment also provides another method for solving the optimal solution for distribution network grid optimization. This method directly calculates the optimal solution for distribution network grid optimization based on the quantum Hamiltonian function. The process is as follows: Figure 3 As shown, the detailed steps are:

[0102] B1. Initialize the magnetic field strength Γ=Γ0 in the quantum Hamiltonian function and the initial temperature T=T o , the number of inner loop iterations K and the outer loop cooling coefficient α, and the lower limit of the magnetic field intensity is set to Γ min .

[0103] B2. Calculate the initial solution and calculate the objective function value F corresponding to the initial solution. Take the objective function value F corresponding to the initial solution as the initial global optimal solution best_F, and calculate the initial Hamiltonian function value H and the initial potential energy term value H corresponding to the initial global optimal solution best_F. pot The method of the initial solution is to calculate a solution that satisfies both the grid radiation constraint and the grid connectivity constraint, and the solution is the initial solution.

[0104] B3. Generate a random disturbance, calculate the disturbance solution under the disturbance and calculate the objective function value corresponding to the disturbance solution, and calculate the disturbance Hamiltonian function value H′ and the disturbance potential energy term value H′ based on the objective function value corresponding to the disturbance solution. pot .

[0105] B4. Calculate the Hamiltonian difference between the initial Hamiltonian function value and the perturbed Hamiltonian function value, i.e., ΔH = H′-H; calculate the potential energy difference between the initial potential energy term value and the perturbed potential energy term value, i.e., ΔH pot =H'pot -H pot ;

[0106] B5. If the Hamiltonian difference ΔH and the potential energy difference ΔH pot If both are greater than or equal to 0, the disturbance is not accepted and random disturbance is regenerated.

[0107] B6. If the Hamiltonian difference and the potential energy difference are both less than the preset values, the disturbance is accepted as the local optimal solution and the corresponding objective function value, and it is determined whether the termination condition is met. If not, the iterative calculation continues. The process of determining the termination condition is as follows:

[0108] C1. Determine whether the number of inner loop iterations has reached the maximum number of iterations. If not, continue iterative calculation.

[0109] C2. If the number of inner loops reaches the maximum number of iterations, the cooling operation T = T0*α and the magnetic field attenuation operation Γ = Γ0*α are performed until the magnetic field strength is less than the lower limit of the magnetic field strength Γ min .

[0110] B7. If satisfied, the current local optimal solution at the time of algorithm termination is taken as the global optimal solution, and the global optimal solution and the corresponding objective function value are output. The global optimal solution is the optimal solution for distribution network optimization.

[0111] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A distribution network optimization method based on quantum annealing algorithm, characterized in that: The method includes: Construct the objective function of distribution network optimization based on actual distribution network data and set constraints; Obtaining a quantum Hamiltonian function based on the objective function and the constraints; Based on the quantum Hamiltonian function, a quadratic unconstrained binary optimization objective function for distribution network optimization is obtained; Solving the optimal solution for distribution network optimization based on the quantum Hamiltonian function or the quadratic unconstrained binary optimization objective function; Generate distribution network optimization strategy based on the optimal solution.

2. A distribution network optimization method based on quantum annealing algorithm according to claim 1, characterized in that: The objective function is determined by the line investment and depreciation cost and the line power loss cost, and its expression is: Where f(x) represents the objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C0 represents the unit price of the line; τ max Indicates the time of maximum load of the distribution system; ΔP i represents the power loss of branch i; ΔI represents the total value of all branch current exceeding the limit; C1 is the branch current exceeding the limit penalty coefficient; ΔU is the total value of all load point voltage exceeding the limit; C2 is the load point voltage exceeding the limit penalty coefficient.

3. A distribution network optimization method based on quantum annealing algorithm according to claim 2, characterized in that: The constraints include branch current constraints, load point current constraints, grid radiation constraints and grid connectivity constraints, wherein the branch current constraint is that the current flowing through the branch does not exceed the maximum current allowed to pass through the branch; the load point current constraint is greater than the load point voltage lower limit and less than the load point voltage lower limit; the grid radiation constraint is that the path from the power source point to any load point is unique and has no loops; the grid connectivity constraint is that the power source point is connected to any load point.

4. A distribution network optimization method based on quantum annealing algorithm according to claim 3, characterized in that: The method for obtaining the quantum Hamiltonian function is to rewrite the objective function according to the grid radiation constraint and the grid connectivity constraint to obtain the quantum Hamiltonian function. The expression of the quantum Hamiltonian function is: Where F(x) represents the quantum Hamiltonian function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C0 represents the unit price of the line; τ max Indicates the time of maximum load of the distribution system; ΔP i represents the power loss of branch i; ΔI represents the total value of all branch current exceeding the limit; C1 is the branch current exceeding the limit penalty coefficient; ΔU is the total value of all load point voltage exceeding the limit; C2 is the load point voltage exceeding the limit penalty coefficient; C3 represents the penalty coefficient of the grid radiation constraint; C4 represents the penalty coefficient of the grid connectivity constraint; C5 represents the no-loop penalty coefficient; L3 represents the penalty function of the grid radiation constraint; L4 represents the penalty function of the grid connectivity constraint; L5 represents the no-loop penalty function; Γ(t) represents the magnetic field strength; represents the potential energy term of the Hamiltonian function.

5. A distribution network optimization method based on quantum annealing algorithm according to claim 4, characterized in that: The method for obtaining the secondary unconstrained binary optimization objective function of the distribution network grid optimization is to extract the line power loss cost calculation item based on the quantum Hamiltonian function, that is, The branch current constraint calculation term, i.e., C1|ΔI|, and the load point current constraint calculation term, i.e., C2|ΔU|, are used as power flow constraints. The line investment and depreciation cost calculation terms based on the quantum Hamiltonian function, i.e., The grid radiation constraint calculation item, namely C3L3+C4L4, and the grid connectivity constraint calculation item, namely C5L5, generate a quadratic unconstrained binary optimization objective function, the expression of which is: Where F'(x) represents the quadratic unconstrained binary optimization objective function; i represents the branch; n represents the total number of branches; γ i represents the capital recovery rate of branch i; represents the annual equipment maintenance rate of the i-th branch; x i represents the decision variable of the i-th branch, that is, when x i =1 means branch i has been built. i =0 means branch i is not built; a i represents the cost of investment and construction per kilometer of branch i; l i represents the construction length of branch i; n represents the total number of branches; C3 represents the penalty coefficient of the grid radiation constraint; C4 represents the penalty coefficient of the grid connectivity constraint; C5 represents the no-loop penalty coefficient; L3 represents the penalty function of the grid radiation constraint; L4 represents the penalty function of the grid connectivity constraint; L4 represents the no-loop penalty function.

6. A distribution network optimization method based on quantum annealing algorithm according to claim 4, characterized in that: The method for solving the optimal solution for distribution network optimization based on the quantum Hamiltonian function is: Initialize the magnetic field strength in the quantum Hamiltonian function, the initial temperature of the algorithm, the number of inner loop iterations, and the outer loop cooling coefficient; Calculate the initial solution and calculate the objective function value corresponding to the initial solution, take the objective function value corresponding to the initial solution as the initial global optimal solution, and calculate the initial Hamiltonian function value and initial potential energy term value corresponding to the initial global optimal solution; Generate a random disturbance, calculate the disturbance solution under the disturbance and calculate the objective function value corresponding to the disturbance solution, and calculate the disturbance Hamiltonian function value and the disturbance potential energy term value based on the objective function value corresponding to the disturbance solution; Calculate the Hamiltonian difference between the initial Hamiltonian function value and the perturbed Hamiltonian function value, and calculate the potential energy difference between the initial potential energy term value and the perturbed potential energy term value; If the Hamiltonian difference and the potential energy difference are both greater than or equal to the preset value, the disturbance is not accepted and a random disturbance is regenerated; If the Hamiltonian difference and the potential energy difference are both less than the preset values, the disturbance is accepted as the local optimal solution and the corresponding objective function value, and it is determined whether the termination condition is met. If not, the iterative calculation is continued; If it is satisfied, the current local optimal solution at the time of algorithm termination is taken as the global optimal solution, and the global optimal solution and the corresponding objective function value are output. The global optimal solution is the optimal solution for distribution network optimization.

7. A distribution network optimization method based on quantum annealing algorithm according to claim 6, characterized in that: The default value is 0.

8. A distribution network optimization method based on quantum annealing algorithm according to claim 6, characterized in that: The termination conditions include: the number of inner loop iterations reaches the maximum number of iterations and the magnetic field strength in the outer loop is less than a preset magnetic field strength.

9. A distribution network optimization method based on quantum annealing algorithm according to claim 5, characterized in that: The method for solving the optimal solution of distribution network optimization based on the quadratic unconstrained binary optimization objective function is: Calculate the initial solution and use power flow to calculate the line power loss cost value corresponding to the initial solution; Generate a quantum Hamiltonian function corresponding to an initial solution, and generate a corresponding quadratic unconstrained binary optimization objective function based on the quantum Hamiltonian function corresponding to the initial solution; Embed the corresponding quadratic unconstrained binary optimization objective function into a D-Wave quantum computer to solve the current optimal solution; Calculate all power flow constraints except the line power loss cost corresponding to the current optimal solution. If all constraints are satisfied, take the current optimal solution as the optimal solution for distribution network optimization. Calculate the line investment and depreciation cost based on the optimal solution for distribution network optimization. Calculate the total cost value in combination with the line power loss cost value of the optimal solution for distribution network optimization. Output the optimal solution for distribution network optimization and the corresponding total cost value. If not, regenerate the solution and repeat the above calculation until the current optimal solution satisfies all power flow constraints.

10. A distribution network optimization method based on quantum annealing algorithm according to claim 6 or claim 9, characterized in that: The method for obtaining the initial solution is to calculate a solution that satisfies both the grid radiation constraint and the grid connectivity constraint, and the solution is the initial solution.

Images