Alternating current and direct current hybrid power distribution network strategy adaptive optimization scheduling method, system, equipment and medium

Abstract

The invention discloses an AC / DC hybrid power distribution network strategy adaptive optimization scheduling method, system, device and medium, and the method comprises the steps: obtaining flexible load power and energy storage system power for constructing an initial solution set; inputting the initial solution set into an optimal scheduling model, and outputting a compromised optimal solution set of the dominating condition; the optimization scheduling model comprises a cost optimization objective function and constraint conditions; evolutionary strategies with different characteristics are set, and solutions in the set are used as reference solutions in the evolutionary strategies to accelerate the optimization process; and updating the initial solution set, carrying out iterative optimization until convergence, randomly selecting optimal flexible load power and energy storage system power from the final compromised optimal solution set, and applying the optimal flexible load power and energy storage system power to the AC / DC hybrid power distribution network. Through an evolution strategy adaptive mechanism, roulette probability feedback and stack score records are combined, the optimization convergence time is greatly shortened, and meanwhile, an AC / DC hybrid power distribution network scene is deeply adapted.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for power systems, and in particular to an adaptive optimization scheduling method, system, equipment, and medium for AC / DC hybrid distribution networks. Background Technology

[0002] The main backgrounds for the development of AC / DC hybrid distribution networks include the following aspects: (1) Rapid development of renewable energy: With the widespread application of renewable energy such as wind and solar power, especially the rise of distributed generation, the power system needs more flexible technologies to connect to and manage these highly volatile power sources. (2) Popularization of electric vehicles: The rapid growth of electric vehicles has put forward new requirements for the power system, especially in terms of charging infrastructure construction and grid load management. Promotion of smart grid technology: The modern power system is transforming towards intelligence and digitalization. AC / DC hybrid distribution networks can better support the functions of smart grids, such as demand response and energy management. Improve energy efficiency and reduce losses: Traditional AC distribution systems have high energy losses in some cases, while DC distribution systems can improve energy efficiency in some application scenarios.

[0003] Optimized scheduling is a crucial aspect of fully leveraging the advantages of AC / DC distribution networks. Reasonable optimized scheduling methods can improve system economy, enhance supply-demand balance, promote renewable energy utilization, and improve system flexibility and adaptability. Commonly used methods in engineering include: ① integer programming; ② particle swarm optimization; ③ hybrid integer programming for carbon flow optimization. However, since the optimization problem of AC / DC hybrid distribution networks is a typical multi-objective optimization problem, conflicts may exist between different objectives, and the computational burden is substantial. How to design new optimized scheduling methods for AC / DC hybrid distribution networks to achieve efficient and high-performance optimized scheduling is an urgent problem to be solved. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an adaptive optimization scheduling method and system for AC / DC hybrid distribution networks to solve the current...

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an adaptive optimization scheduling method for AC / DC hybrid distribution network strategy, comprising: obtaining the power of flexible loads and the power of energy storage systems to construct an initial solution set;

[0008] The initial solution set is input into the optimization scheduling model, which outputs a compromise optimal solution set for the dominance situation; the optimization scheduling model includes a cost optimization objective function and constraints.

[0009] Different evolutionary strategies with different characteristics are set, and the solutions in the set are used as reference solutions in the evolutionary strategies to accelerate the optimization process;

[0010] The initial solution set is updated, and iterative optimization is performed until convergence. The optimal flexible load power and energy storage system power are randomly selected from the final compromise optimal solution set and applied to the AC / DC hybrid distribution network.

[0011] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution networks described in this invention, the initial solution set is constructed, including:

[0012] An initial solution set X is constructed by generating a solution set through random initialization. t The total number of solutions is assumed to be N;

[0013] Each solution includes two control parameters: the power of the flexible load and the power of the energy storage system.

[0014] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution network described in this invention, wherein: the initial solution set is input into the optimization scheduling model, and the optimization scheduling model includes an objective function and constraints;

[0015] The objective function includes minimizing the costs of electricity purchase, wind and solar curtailment penalties, grid loss penalties, and flexible load dispatching.

[0016] The constraints include power flow constraints, security constraints, and energy storage constraints.

[0017] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution networks described in this invention, the initial solution set is input into the optimization scheduling model, and the compromise optimal solution set of the dominance situation is output, including:

[0018] Each element in the initial solution set is input into the objective function of the optimization scheduling model, and in each iteration, the cost of electricity purchase, the cost of wind and solar curtailment penalty, the cost of grid loss penalty, and the cost of flexible load scheduling are calculated based on the current solution set.

[0019] Compare each solution pairwise and define a solution. Dominate another solution All objective function values ​​must be no worse than And at least one objective is better;

[0020] The dominance of all solutions is statistically analyzed, and the solutions that are not dominated by any solution are stored in the compromise optimal solution set and output.

[0021] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution networks described in this invention, the method includes: setting evolution strategies with different characteristics, and using solutions in the set as reference solutions in the evolution strategies to accelerate the optimization process, including:

[0022] For each strategy, a first-in-first-out stack structure is established to record the scores obtained by each strategy in successfully generating better experimental individuals during the learning cycles of the previous c generations of the current evolutionary generation.

[0023] Initialize all stack units so that the strategy scores are the same;

[0024] Randomly select solutions from the set and input them into the stack structure;

[0025] At the end of each iteration, the score of the current differential evolution strategy is pushed to the top of the stack, the remaining units are moved down, and the oldest score record is removed from the bottom of the stack.

[0026] The beneficial effects of this preferred technical solution are: by setting an evolutionary strategy, it can overcome the shortcomings of mixed integer programming, particle swarm optimization and other methods in high-dimensional multi-objective optimization, such as being prone to getting trapped in local optima and having slow convergence speed.

[0027] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution networks described in this invention, the method further includes: setting evolution strategies with different characteristics, using solutions in the set as reference solutions in the evolution strategies to accelerate the optimization process; and further includes:

[0028] After all strategies have been scored, update the probability of the score being selected in the next iteration.

[0029] During the iteration process, when each objective solution mutates, a roulette wheel selection mechanism is used to randomly select a strategy based on the corresponding probability.

[0030] The beneficial effects of this preferred technical solution are: based on the historical score of the strategy, the probability of selecting high-quality strategies is increased, which can balance randomness and guidance.

[0031] As a preferred embodiment of the adaptive optimization scheduling method for AC / DC hybrid distribution networks described in this invention, the step of updating the initial solution set and performing iterative optimization until convergence includes:

[0032] For N preset solutions, each preset solution is replaced with the corresponding mutated solution generated by the evolutionary strategy in the current generation;

[0033] After updating the solution set, the initial solution calculation and the output of the compromise optimal solution set for the dominance situation are performed repeatedly until the preset maximum number of iterations is reached;

[0034] From the final compromise optimal solution set, a set of optimal parameter combinations is randomly selected and input into the AC / DC hybrid distribution network control system to drive the equipment to execute the scheduling strategy.

[0035] Secondly, the present invention provides an adaptive optimization scheduling system for AC / DC hybrid distribution networks, comprising:

[0036] The acquisition module is used to acquire the power of the flexible load and the power of the energy storage system, and is used to construct the initial solution set;

[0037] The module is used to input the initial solution set into the optimization scheduling model and output the compromise optimal solution set of the dominance situation; the optimization scheduling model includes a cost optimization objective function and constraints.

[0038] The optimization module is used to set evolution strategies with different characteristics, and uses the solutions in the set as reference solutions in the evolution strategy to accelerate the optimization process;

[0039] The update module is used to update the initial solution set, perform iterative optimization until convergence, and randomly select the optimal flexible load power and energy storage system power from the final compromise optimal solution set for application in AC / DC hybrid distribution networks.

[0040] Thirdly, the present invention provides a computer device, comprising:

[0041] Memory and processor;

[0042] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy are implemented.

[0043] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the AC / DC hybrid distribution network strategy adaptive optimization scheduling method.

[0044] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention aims to minimize the sum of electricity purchase costs, wind and solar curtailment penalty costs, grid loss penalty costs, and flexible load dispatch costs, achieving efficient optimization of the objective. Through an evolutionary strategy adaptive mechanism, combined with roulette wheel probability feedback and stacked score recording, the optimization convergence time is significantly shortened, while also being deeply adapted to AC / DC hybrid distribution network scenarios. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the overall process of the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy according to an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of a typical AC / DC hybrid distribution network in the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy according to an embodiment of the present invention.

[0048] Figure 3 This is a schematic diagram of the bus voltage balance control method of the photovoltaic-storage-DC-flexible system in the adaptive optimization scheduling method of the AC / DC hybrid distribution network strategy according to an embodiment of the present invention.

[0049] Figure 4 This is a schematic diagram of the stack operation process in the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy according to an embodiment of the present invention.

[0050] Figure 5 This is a schematic diagram of the wheel probability selection mechanism in the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy according to an embodiment of the present invention. Detailed Implementation

[0051] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0052] Example 1, referring to Figures 1-2 As an embodiment of the present invention, an adaptive optimization scheduling method for AC / DC hybrid distribution networks is provided, comprising:

[0053] S100: Obtain the power of the flexible load and the power of the energy storage system to construct the initial solution set;

[0054] S200: Input the initial solution set into the optimization scheduling model and output the compromise optimal solution set of the dominance situation; the optimization scheduling model includes a cost optimization objective function and constraints;

[0055] S300: Set evolution strategies with different characteristics, and use the solutions in the set as reference solutions in the evolution strategy to accelerate the optimization process;

[0056] S400: Update the initial solution set, perform iterative optimization until convergence, and randomly select the optimal flexible load power and energy storage system power from the final compromise optimal solution set for application in AC / DC hybrid distribution networks.

[0057] It should be noted that, Figure 2 The diagram shows a typical AC / DC hybrid distribution network, which is usually composed of new energy power generation equipment such as wind power and photovoltaic, power electronic converters such as AC / DC and DC / DC, energy storage, DC loads, AC loads, etc. Its advantages are mainly: (1) It is compatible with multiple power sources. The AC / DC hybrid distribution network can connect to multiple power sources, including traditional power generation, renewable energy and energy storage systems, which enhances the flexibility and adaptability of the power system. (2) It supports distributed generation. This distribution network structure can better support the access of distributed generation, realize the local consumption of electricity, and improve the stability of the system. (3) It adapts to emerging loads. The AC / DC hybrid distribution network can effectively meet the needs of emerging loads such as electric vehicle charging and data centers, and has stronger adaptability to DC loads. (4) It reduces system complexity. By using DC distribution, the design and control of power electronic equipment can be simplified, the number of conversion equipment can be reduced, thereby reducing the complexity and cost of the system. (5) It enhances the system's anti-disturbance capability. The AC / DC hybrid distribution network can improve the system's response capability to sudden events (such as faults, load fluctuations, etc.), and enhance the stability and reliability of the power system.

[0058] However, existing algorithms such as integer programming and particle swarm optimization suffer from several drawbacks when solving multi-objective problems in AC / DC hybrid distribution networks. Firstly, they require layer-by-layer linearization of the power flow equations followed by distributed coordinated iterative solutions, leading to an exponential increase in computational complexity. Secondly, they rely on swarm iteration and mesh partitioning, resulting in slow convergence when alternately solving two-layer models, with optimization of 100-node networks often exceeding one hour. Furthermore, they lack dynamic adjustment mechanisms; typical integer programming algorithms rely on fixed solvers and cannot switch optimization strategies based on solution space characteristics, making them ill-suited for complex power grid scenarios.

[0059] To address the aforementioned main issues, steps S100-S400 employ an adaptive evolution strategy to improve optimization efficiency. An adaptive evolution probability coefficient mechanism is adopted to minimize the sum of electricity purchase costs, wind and solar curtailment penalty costs, grid loss penalty costs, and flexible load dispatch costs, thereby achieving efficient optimization of the objectives.

[0060] Example 2, refer to Figures 1-5 As an embodiment of the present invention, based on the above embodiment, an adaptive optimization scheduling method for AC / DC hybrid distribution network strategy is provided.

[0061] In the embodiments of this application, such as Figure 3 As shown, step S100 obtains the power of the flexible load and the power of the energy storage system to construct an initial solution set, including steps A1-A2:

[0062] A1: Generate a solution set by random initialization and construct an initial solution set X. t The total number of solutions is assumed to be N;

[0063] A2: Each solution includes two control parameters: flexible load power and energy storage system power.

[0064] Specifically, Let t be the evolutionary algebra. For the initial solution set t = 0, any n∈[1,N], where For the initial solution of flexible load power, This represents the initial power solution for the energy storage system, with units of kW.

[0065] In an alternative implementation, the initial solution set in step S100 can also be generated by historical data. Compared with random initialization, this can avoid the generation of invalid solutions and improve the quality of the initial solution.

[0066] In another alternative implementation, the initial solution set in step S100 can also be constructed by sampling the constraint boundaries, for example: identifying the power flow constraint safety threshold, the energy storage charge and discharge limit, and the voltage stability margin, and generating characteristic solutions at the constraint boundaries.

[0067] In this embodiment of the application, step S200, which involves inputting the initial solution set into the optimization scheduling model, includes an objective function and constraints.

[0068] The objective function includes minimizing the costs of electricity purchase, wind and solar curtailment penalties, grid loss penalties, and flexible load dispatching.

[0069] For example, the objective function is shown in formula (1):

[0070] min[f1(x)=C pur (x1,x2); f2(x) = C DG (x1,x2); f3(x) = C loss (x1,x2);

[0071] f4(x)=C load (x1,:x2)] (1)

[0072] Where x1 and x2 are the control variables, flexible load power and energy storage system, respectively; C pur (x1, x2) represents the electricity purchase cost obtained from this control variable; C DG(x1, x2) represents the cost of wind and solar power curtailment penalties; C DG (x1, x2) represents the network loss penalty cost; C load (x1,:x2) represents the cost of flexible load scheduling.

[0073] The constraints include power flow constraints, security constraints, and energy storage constraints.

[0074] For example, the constraints are shown in formula (2):

[0075] subject[g1(x)=S cl (x1,x2); g2(x) = S aq (x1,x2); g3(x) = S cn (x1,x2)] (2)

[0076] Among them, S cl (x1, x2) represents the power flow constraint; S aq (x1, x2) are safety constraints; S cn (x1, x2) represents the energy storage constraints.

[0077] In an optional implementation, the objective function in step S200 may further include minimizing carbon emission costs, which can be combined with... Figure 2 Carbon emissions from new energy equipment on the DC bus can be measured in real time.

[0078] In another optional implementation, the constraints in step S200 may also include power quality constraints, which can trigger a forced correction of power quality constraints when the energy storage power changes abruptly.

[0079] In this embodiment of the application, step S200 inputs the initial solution set into the optimization scheduling model and outputs the compromise optimal solution set of the dominance situation, including the following steps B1-B3:

[0080] B1: Input each element in the initial solution set into the objective function of the optimization scheduling model, and in each iteration, calculate the objective function value based on the current solution set to obtain the electricity purchase cost, wind and solar curtailment penalty cost, grid loss penalty cost, and flexible load scheduling cost;

[0081] For example, will Substituting each element of t=0 into formula (1), we obtain...

[0082] B2: Compare each solution pairwise and define a solution. Dominate another solution All objective function values ​​must be no worse than And at least one objective is better;

[0083] For example, define Dominate Must meet:

[0084]

[0085] If in If not dominated by any other solution, then This is known as the optimal compromise solution.

[0086] B3: Calculate the dominance of all solutions, and store the solutions that are not dominated by any solution into the compromise optimal solution set for output;

[0087] For example, according to the above definition, for all The dominance of n = [1, N] is statistically analyzed; and all compromise optimal solutions are put into set O, which is a set that stores the compromise optimal solutions calculated during the evolution process.

[0088] In this embodiment of the application, step S300 sets evolutionary strategies with different characteristics. Specifically, six evolutionary strategies can be defined, which can be represented as follows:

[0089] ①

[0090] ②

[0091] ③

[0092] ④

[0093] ⑤

[0094] ⑥

[0095] in, For each generation t, from the solution set X t Three solutions are randomly selected from the given data. Each solution includes a flexible load power value and an energy storage system power value. F is a proportionality coefficient in the range of 0-1. Let t be any compromise optimal solution chosen from the solution set.

[0096] It should be noted that each of the six strategies has different characteristics. Strategy ① has strong global exploration capabilities; Strategy ② uses the compromise optimal solution as the mutation basis, and the difference term provides directional perturbation; Strategy ③ has a double difference term that can significantly amplify the perturbation amplitude; Strategy ④ can explore the neighborhood of high-quality solutions in depth; Strategy ⑤ develops elite directions for the F1 term and maintains random exploration for the F2 term. Figure 4The probabilistic mechanism prioritizes this strategy to handle complex conflict scenarios; strategy ⑥ expands gradually with the current solution as the base address, and double random difference avoids large deviations from the feasible region.

[0097] In one alternative implementation, the ratio of exploratory strategies (①③), developmental strategies (②④), and hybrid strategies (⑤⑥) is dynamically balanced by the probability of roulette wheel betting, and a corresponding variant solution is obtained by selecting one of these strategies according to a specific probability.

[0098] In another alternative implementation, the hybrid strategy (⑤⑥) can be used as a transitional solution to improve robustness.

[0099] In this embodiment of the application, step S300 sets evolutionary strategies with different characteristics, and uses the solutions in the set as reference solutions in the evolutionary strategies to accelerate the optimization process, including steps C1-C4:

[0100] C1: For each strategy, establish a first-in-first-out (FIFO) stack structure to record the scores obtained by each strategy in successfully generating better experimental individuals during the learning cycles of the previous c generations of the current evolutionary generation.

[0101] For example, based on the evolutionary strategy, the number of stack structures can be set to K, i.e., K=6.

[0102] C2: Initialize all stack units to make the strategy scores the same;

[0103] C3: Randomly select solutions from the set and input them into the stack structure;

[0104] C4: At the end of each iteration, the score of the current differential evolution strategy is pushed to the top of the stack, the remaining units are moved down, and the oldest score record is removed from the bottom of the stack.

[0105] Specifically, the above steps can be achieved through... This represents the score of the j-th strategy during the t-th generation evolution process.

[0106]

[0107] In the formula, M represents the number of solutions using the k-th strategy. Let represent the score of the i-th solution during generation t. If the i-th solution is the optimal compromise solution, then... otherwise

[0108] In this embodiment of the application, step S300 sets evolutionary strategies with different characteristics, uses the solutions in the set as reference solutions in the evolutionary strategies to accelerate the optimization process, and also includes the following steps D1-D2:

[0109] D1: After all strategies have been scored, update the probability of the score being selected in the next iteration;

[0110] D2: During the iteration process, when each objective solution mutates, a roulette wheel selection mechanism is used to randomly select a strategy based on the corresponding probability.

[0111] Specifically, after the scores of all K strategies are calculated, their probabilities of being selected in the next generation of evolution are updated using formula (5).

[0112]

[0113] Figure 5 A schematic diagram of the roulette wheel selection mechanism is shown. Each strategy occupies a corresponding angle area on the roulette wheel according to its probability value. The higher the probability value, the larger the angle it occupies, and thus the greater the probability of being selected.

[0114] In an alternative implementation, in step D2, in addition to using the roulette wheel selection mechanism, a priority queue selection mechanism can also be used. That is, a priority queue is dynamically constructed based on the strategy score, and high-value strategies are called first each time a selection is made. At the same time, randomness is retained to avoid premature convergence, which can reduce the frequency of inefficient strategy calls.

[0115] In another alternative implementation, in addition to using the roulette wheel selection mechanism, a conflict scenario adaptive selection mechanism can also be used. For example, the dimension of the D1 step score calculation can be expanded, a conflict intensity factor can be added, and the D2 selection response can be made to the multi-objective conflict state. The strategy selection weight can be dynamically adjusted based on the conflict intensity of the objective function, and a targeted strategy library can be enabled in the strong conflict scenario.

[0116] It should be noted that the preferred roulette wheel selection mechanism in this application is because it can be directly used for probability calculation, belongs to the full-strategy roulette wheel, is more comprehensive, has low complexity, and has a high degree of adaptability.

[0117] In this embodiment of the application, the step S400 of updating the initial solution set and performing iterative optimization until convergence includes the following steps E1-E3:

[0118] E1: For N preset solutions, replace each preset solution with the corresponding mutated solution generated by the evolution strategy in the current generation;

[0119] E2: After updating the solution set, repeatedly execute the steps of calculating the initial solution and outputting the compromise optimal solution set of the dominance situation until the preset maximum number of iterations is reached;

[0120] E3: Randomly select a set of optimal parameter combinations from the final compromise optimal solution set, input them into the AC / DC hybrid distribution network control system, and drive the equipment to execute the scheduling strategy.

[0121] For example, for all i∈[1,N], let Then the initial solution calculation is performed again until t = t. max Then all compromise optimal solutions can be obtained, and one of the compromise optimal solutions can be randomly selected as the flexible load power. Energy storage system power It also applies to AC / DC hybrid distribution networks.

[0122] It should be noted that the above evolutionary strategy mechanism can also break through the rigidity of fixed strategies and adapt to complex fluctuation scenarios of AC / DC distribution networks through a dynamic strategy library, such as sudden changes in wind and solar power output and sharp increases in load; the six strategies focus on global exploration, local development or hybrid optimization, respectively, covering the optimization needs throughout the entire cycle.

[0123] In summary, the AC / DC hybrid distribution network strategy adaptive optimization scheduling method of the present invention uses flexible load power and energy storage system power as control variables, and minimizes the sum of power purchase cost, wind and solar curtailment penalty cost, network loss penalty cost and flexible load scheduling cost as the optimization objective. By constructing an evolutionary strategy adaptive mechanism, it realizes multi-objective optimization scheduling of AC / DC hybrid distribution networks.

[0124] Example 3 illustrates a schematic scheme for an adaptive optimization scheduling method for AC / DC hybrid distribution networks. It should be noted that the technical solution of this adaptive optimization scheduling system for AC / DC hybrid distribution networks is based on the same concept as the technical solution of the aforementioned adaptive optimization scheduling method for AC / DC hybrid distribution networks. Details not described in detail in this example can be found in the description of the technical solution of the aforementioned adaptive optimization scheduling method for AC / DC hybrid distribution networks.

[0125] This embodiment also provides another AC / DC hybrid distribution network strategy adaptive optimization scheduling system, including:

[0126] The acquisition module is used to acquire the power of the flexible load and the power of the energy storage system, and is used to construct the initial solution set;

[0127] The module is used to input the initial solution set into the optimization scheduling model and output the compromise optimal solution set of the dominance situation; the optimization scheduling model includes a cost optimization objective function and constraints.

[0128] The optimization module is used to set evolution strategies with different characteristics, and uses the solutions in the set as reference solutions in the evolution strategy to accelerate the optimization process;

[0129] The update module is used to update the initial solution set, perform iterative optimization until convergence, and randomly select the optimal flexible load power and energy storage system power from the final compromise optimal solution set for application in AC / DC hybrid distribution networks.

[0130] This embodiment also provides a computer device suitable for adaptive optimization scheduling of AC / DC hybrid distribution network strategies, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the adaptive optimization scheduling method for AC / DC hybrid distribution network strategies proposed in the above embodiments.

[0131] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy proposed in the above embodiments.

[0132] The storage medium proposed in this embodiment and the adaptive optimization scheduling method for implementing AC / DC hybrid distribution network strategy proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0133] Based on the above description of the implementation methods, those skilled in the art will clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0134] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An adaptive optimization scheduling method for AC / DC hybrid distribution networks, characterized in that, include: The power of the flexible load and the power of the energy storage system are obtained to construct the initial solution set; The initial solution set is input into the optimization scheduling model, and the output is the set of compromise optimal solutions for the dominance situation; The optimized scheduling model includes a cost optimization objective function and constraints. Different evolutionary strategies with different characteristics are set, and the solutions in the set are used as reference solutions in the evolutionary strategies to accelerate the optimization process; The initial solution set is updated, and iterative optimization is performed until convergence. The optimal flexible load power and energy storage system power are randomly selected from the final compromise optimal solution set and applied to the AC / DC hybrid distribution network.

2. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 1, characterized in that, Construct the initial solution set, including: An initial solution set X is constructed by generating a solution set through random initialization. t The total number of solutions is assumed to be N; Each solution includes two control parameters: the power of the flexible load and the power of the energy storage system.

3. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 2, characterized in that, The initial solution set is input into the optimization scheduling model, wherein the optimization scheduling model includes an objective function and constraints; The objective function includes minimizing the costs of electricity purchase, wind and solar curtailment penalties, grid loss penalties, and flexible load dispatching. The constraints include power flow constraints, security constraints, and energy storage constraints.

4. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 3, characterized in that, The initial solution set is input into the optimization scheduling model, which outputs a set of compromise optimal solutions for the dominance situation, including: Each element in the initial solution set is input into the objective function of the optimization scheduling model, and in each iteration, the cost of electricity purchase, the cost of wind and solar curtailment penalty, the cost of grid loss penalty, and the cost of flexible load scheduling are calculated based on the current solution set. Compare each solution pairwise and define a solution. Dominate another solution All objective function values ​​must be no worse than And at least one objective is better; The dominance of all solutions is statistically analyzed, and the solutions that are not dominated by any solution are stored in the compromise optimal solution set and output.

5. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 4, characterized in that, Different evolutionary strategies with different characteristics are defined, and solutions in the set are used as reference solutions in the evolutionary strategies to accelerate the optimization process, including: For each strategy, a first-in-first-out stack structure is established to record the scores obtained by each strategy in successfully generating better experimental individuals during the learning cycles of the previous c generations of the current evolutionary generation. Initialize all stack units so that the strategy scores are the same; Randomly select solutions from the set and input them into the stack structure; At the end of each iteration, the score of the current differential evolution strategy is pushed to the top of the stack, the remaining units are moved down, and the oldest score record is removed from the bottom of the stack.

6. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 5, characterized in that, Different evolutionary strategies with different characteristics are defined, and the solutions in the set are used as reference solutions in the evolutionary strategies to accelerate the optimization process. This also includes: After all strategies have been scored, update the probability of the score being selected in the next iteration. During the iteration process, when each objective solution mutates, a roulette wheel selection mechanism is used to randomly select a strategy based on the corresponding probability.

7. The adaptive optimization scheduling method for AC / DC hybrid distribution network as described in claim 6, characterized in that, The process of updating the initial solution set and iteratively optimizing it until convergence includes: For N preset solutions, each preset solution is replaced with the corresponding mutated solution generated by the evolutionary strategy in the current generation; After updating the solution set, the initial solution calculation and the output of the compromise optimal solution set for the dominance situation are performed repeatedly until the preset maximum number of iterations is reached; From the final compromise optimal solution set, a set of optimal parameter combinations is randomly selected and input into the AC / DC hybrid distribution network control system to drive the equipment to execute the scheduling strategy.

8. An adaptive optimization scheduling system for AC / DC hybrid distribution networks, employing the method described in any one of claims 1-7, characterized in that, include: The acquisition module is used to acquire the power of the flexible load and the power of the energy storage system, and is used to construct the initial solution set; The module is used to input the initial solution set into the optimization scheduling model and output the compromise optimal solution set of the dominance situation; the optimization scheduling model includes a cost optimization objective function and constraints. The optimization module is used to set evolution strategies with different characteristics, and uses the solutions in the set as reference solutions in the evolution strategy to accelerate the optimization process; The update module is used to update the initial solution set, perform iterative optimization until convergence, and randomly select the optimal flexible load power and energy storage system power from the final compromise optimal solution set for application in AC / DC hybrid distribution networks.

9. A computer device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the adaptive optimization scheduling method for AC / DC hybrid distribution network strategy as described in any one of claims 1 to 7.

Images