A medium-voltage power distribution network flexible interconnection backbone network architecture planning method, system, device, medium and product

Abstract

The application discloses a medium-voltage distribution network flexible interconnection backbone network framework planning method, system, equipment, medium and product, relates to the flexible interconnection device configuration technical field, obtains the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, carries out enumeration preliminary screening according to the candidate decision set, obtains a plurality of candidate schemes, generates the corresponding initial operation strategy population according to each candidate scheme respectively, iteratively optimizes the initial operation strategy population corresponding to each candidate scheme by using the pre-trained prediction network and electrical quantity data, obtains a plurality of target operation strategies and a plurality of target prediction data, filters each candidate scheme and each target operation strategy according to each target prediction data, and obtains a flexible interconnection backbone network framework planning scheme. The technical problem that the existing network framework planning method is prone to low equipment utilization and reduces the overall operation efficiency of the distribution network is solved.

Description

Technical Field

[0001] This invention relates to the field of flexible interconnection device configuration technology, and in particular to a planning method, system, equipment, medium and product for a flexible interconnection backbone network of a medium-voltage distribution network. Background Technology

[0002] With the continuous advancement of new power system construction, distribution networks are rapidly transforming from traditional radial structures to flexible interconnected forms. Power electronics technology, with flexible interconnection devices (FIDs) at its core, enables flexible power exchange between feeders, demonstrating significant application value in improving renewable energy absorption, optimizing voltage quality, and balancing load distribution. Actual engineering verification shows that in distribution network areas with scientifically configured flexible interconnection devices, renewable energy absorption rates can increase by more than 15%, system network losses can decrease by 8%–12%, and power supply reliability indicators SAIDI and SAIFI can improve by 20%–30%, becoming a key technological support for promoting the flexible and intelligent upgrading of distribution networks.

[0003] Currently, most existing planning methods for flexible interconnected backbone networks of medium-voltage distribution networks adopt a hierarchical optimization framework, which decouples medium- and long-term planning from short-term operation and control. This results in a mismatch between the planning scheme and the actual operating conditions, leading to problems such as low equipment utilization and poor investment efficiency, which reduces the overall operating efficiency of the distribution network. Summary of the Invention

[0004] This invention provides a planning method, system, equipment, medium, and product for a flexible interconnected backbone network of medium-voltage distribution networks. It solves the technical problem that existing planning methods for flexible interconnected backbone networks of medium-voltage distribution networks often adopt a hierarchical optimization framework, which decouples medium- and long-term planning from short-term operation control. This results in a mismatch between the planning scheme and the actual operating conditions, leading to problems such as low equipment utilization and poor investment efficiency, thus reducing the overall operating efficiency of the distribution network.

[0005] The first aspect of this invention provides a method for planning a flexible interconnected backbone network for a medium-voltage distribution network, comprising:

[0006] Obtain the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and perform initial screening based on the candidate decision set to obtain multiple candidate schemes;

[0007] Each candidate scheme generates a corresponding initial operating strategy population, and the pre-trained prediction network and the electrical quantity data are used to iteratively optimize the initial operating strategy population corresponding to each candidate scheme to obtain multiple target operating strategies and multiple target prediction data.

[0008] Based on the predicted data for each target, the candidate schemes and the operational strategies for each target are screened to obtain the corresponding flexible interconnected backbone network planning scheme.

[0009] Optionally, the step of enumerating and initially screening based on the candidate decision set to obtain multiple candidate solutions includes:

[0010] Based on preset quantity and capacity constraints, feasible solutions corresponding to the candidate decision set are generated by enumeration.

[0011] Each of the feasible solutions is input into a preset investment cost function to obtain multiple investment costs;

[0012] When the investment cost is less than or equal to a preset budget threshold, the feasible solution corresponding to the investment cost is determined as a candidate solution.

[0013] By adopting the above technical solution, feasible solutions are generated through enumeration in the early planning stage, combining quantity and capacity constraints. This eliminates unreasonable decision combinations that do not meet equipment configuration limitations from the outset, ensuring that the solutions meet basic engineering implementation conditions. Then, each feasible solution is quantitatively calculated using an investment cost function, and economic screening is performed based on budget thresholds. This achieves dual verification of technical feasibility and economic rationality, preventing invalid solutions from entering subsequent complex optimization processes to reduce computational overhead, and ensuring that the final selected candidate solutions all meet actual engineering constraints. This lays a compliant and efficient foundation for the subsequent collaborative optimization of grid planning and operation strategies.

[0014] Optionally, the step of iteratively optimizing the initial operating strategy population corresponding to each candidate scheme using a pre-trained prediction network and the electrical quantity data to obtain multiple target operating strategies and multiple target prediction data includes:

[0015] Each individual in the population of each initial operating strategy is given a corresponding regional load curve, wherein each individual corresponds to an initial operating strategy.

[0016] The regional load curves, candidate schemes, electrical quantity data, and switch states in the initial operation strategy corresponding to each individual are respectively input into a pre-trained prediction network to obtain multiple prediction data.

[0017] Each of the predicted data is input into a preset fitness function to obtain multiple fitness values;

[0018] Each initial running strategy population is optimized based on its fitness to obtain multiple target running strategies;

[0019] The predicted data corresponding to each of the target operation strategies is used as the target prediction data.

[0020] By adopting the above technical solution, a corresponding regional load curve is generated for each individual in the initial operating strategy population, transforming the abstract operating strategy into quantifiable load data, providing an input basis that fits the actual operating scenario for subsequent predictions. Secondly, the regional load curve, candidate schemes, electrical quantity data, and switch status data are jointly input into a pre-trained prediction network to ensure that the prediction data comprehensively reflects the operating status of the schemes and improves prediction accuracy. Then, the prediction data is quantitatively evaluated using a fitness function, providing a clear basis for judging the merits of the operating strategy population optimization. Subsequently, the population is iteratively optimized based on fitness, gradually eliminating inferior strategies and evolving superior strategies, ultimately obtaining the target operating strategy that fits each candidate scheme. Simultaneously, the prediction data corresponding to the target operating strategy is determined as the target prediction data, ensuring the optimality of the target operating strategy and providing high-precision, high-reference-value quantitative indicators for subsequent scheme selection. This effectively improves the efficiency and rationality of operating strategy optimization, avoids blind optimization, and achieves coordinated adaptation between the operating strategy and the network planning, providing solid support for the final selection of flexible interconnected backbone network planning schemes.

[0021] Optionally, the step of filtering each candidate scheme and each target operation strategy based on each target prediction data to obtain the corresponding flexible interconnection backbone network planning scheme includes:

[0022] Based on preset technical weights, the predicted data of each target are weighted and calculated to obtain multiple technical benefit values;

[0023] The network loss data corresponding to each of the target operation strategies is input into a preset network loss reduction benefit function to obtain multiple network loss reduction benefits;

[0024] The expected power shortage of the preset baseline system is compared with the expected power shortage of the optimized system in each of the target prediction data to obtain multiple first differences;

[0025] Each of the first differences is multiplied by the corresponding unit power outage loss electricity price to obtain multiple power outage loss reduction benefits;

[0026] The benefits of reducing network losses are summed with the corresponding benefits of reducing power outage losses and delaying the return on power grid investment to obtain multiple economic benefit values.

[0027] The investment data of each candidate scheme is obtained, and each investment data is input into a preset equal annual value function to obtain multiple investment cost penalty values;

[0028] Based on preset evaluation weights, the technical benefit value, economic benefit value, investment cost penalty value and carbon reduction benefit value corresponding to each candidate scheme are weighted and calculated to obtain multiple target benefit evaluation values;

[0029] The candidate scheme and target operation strategy corresponding to the maximum value among the various target benefit evaluation values ​​are selected as the corresponding flexible interconnection backbone network planning scheme.

[0030] By adopting the above technical solutions, and through step-by-step quantitative calculations of technical benefits, network loss reduction benefits, power outage loss reduction benefits, and carbon reduction gains, the abstract advantages of the solutions are transformed into precisely comparable quantitative indicators. This avoids the bias of subjective judgment and makes the comparison of the merits of each candidate solution clearer and more objective, ensuring the fairness of the evaluation results. Simultaneously, using preset evaluation weights as a link, multi-dimensional indicators of technology, economy, cost, and environmental protection (carbon reduction) are integrated and weighted, breaking the limitations of single-dimensional evaluation. This ensures that the final selected solution not only meets the requirements of technical feasibility but also possesses economic rationality and environmental compliance, truly achieving the planning goal of "technologically feasible, economically optimal, cost-controllable, and green and low-carbon." Fourth, the entire evaluation process is logically coherent and the steps are clear. From single benefit calculation to comprehensive weighting, from cost penalty to final selection, a closed-loop evaluation system is formed. This not only improves the efficiency and accuracy of scheme selection, but also ensures that the final flexible interconnected backbone network planning scheme can adapt to the actual operation needs of the medium-voltage distribution network, taking into account both practicality and long-term effectiveness. It effectively avoids the problems of disconnect between planning and actual operation and imbalance between benefits and costs, and provides scientific support for the flexible and intelligent development of the medium-voltage distribution network.

[0031] Optionally, the training process of the prediction network is as follows:

[0032] Multiple training network planning data are acquired, and each training network planning data is preprocessed to obtain the corresponding prediction feature set.

[0033] The predicted feature set is input into a preset error backpropagation neural network for training to obtain the corresponding training prediction data;

[0034] Based on a preset loss function, the loss function value of the prediction feature set is determined according to the training prediction data;

[0035] When the loss function value is greater than or equal to a preset loss threshold, the network parameters of the backpropagation neural network are adjusted until the loss function value is less than the loss threshold.

[0036] When the loss function value is less than the loss threshold, a corresponding prediction network is generated.

[0037] Optionally, the isochronous value function is specifically:

[0038] ;

[0039] in, As the penalty value for investment costs, The total investment cost, This is the capital recovery coefficient. Annual operating costs, Let n be the discount rate and n be the project lifespan.

[0040] The second aspect of this invention provides a flexible interconnection backbone network planning system for medium-voltage distribution networks, comprising:

[0041] The acquisition module is used to acquire the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and to perform enumeration and initial screening based on the candidate decision set to obtain multiple candidate schemes;

[0042] The prediction module is used to generate a corresponding initial operation strategy population based on each of the candidate schemes, and to iteratively optimize the initial operation strategy populations corresponding to each of the candidate schemes using a pre-trained prediction network and the electrical quantity data, so as to obtain multiple target operation strategies and multiple target prediction data.

[0043] The filtering module is used to filter each candidate scheme and each target operation strategy based on the prediction data of each target, so as to obtain the corresponding flexible interconnection backbone network planning scheme.

[0044] The third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the medium-voltage distribution network flexible interconnection backbone network planning method as described above.

[0045] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the medium-voltage distribution network flexible interconnection backbone network planning method as described above.

[0046] The fifth aspect of the present invention provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein, when the program instructions are executed by a computer, the computer performs the medium-voltage distribution network flexible interconnection backbone network planning method as described above.

[0047] As can be seen from the above technical solutions, the present invention has the following advantages:

[0048] By acquiring the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and performing initial screening based on the candidate decision set to obtain multiple candidate schemes, a corresponding initial operation strategy population is generated for each candidate scheme. A pre-trained prediction network and electrical quantity data are then used to iteratively optimize the initial operation strategy populations for each candidate scheme, resulting in multiple target operation strategies and multiple target prediction data. Based on the target prediction data, each candidate scheme and each target operation strategy is screened to obtain the corresponding flexible interconnected backbone network planning scheme. This approach overcomes the technical problem that existing medium-voltage distribution network flexible interconnected backbone network planning methods often employ a hierarchical optimization framework, decoupling medium- and long-term planning from short-term operation control. This leads to mismatches between the planned scheme and actual operating conditions, resulting in low equipment utilization, poor investment efficiency, and reduced overall operational efficiency of the distribution network. Compared with traditional planning methods, this approach quickly eliminates invalid schemes that do not meet the constraints through enumeration and initial screening. Then, iterative optimization of the operation strategy is carried out by relying on genetic algorithms, pre-trained prediction networks, and electrical quantity data. This achieves collaborative solution of network planning and operation strategy, ensuring the feasibility of candidate schemes in terms of quantity, capacity, and investment. At the same time, the optimization efficiency and accuracy of the operation strategy are greatly improved by using a high-precision prediction network. Furthermore, the screening of each candidate scheme and each target operation strategy based on the prediction data of each target ensures that the flexible interconnected backbone network planning scheme has technical rationality, economic advantages, and operational reliability. This effectively avoids the problem of planning and operation being disconnected and improves the overall operation efficiency of the distribution network. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0050] Figure 1 This is a flowchart illustrating the steps of a flexible interconnection backbone network planning method for medium-voltage distribution networks provided in Embodiment 1 of the present invention.

[0051] Figure 2 This is a flowchart illustrating the steps of a flexible interconnection backbone network planning method for medium-voltage distribution networks provided in Embodiment 2 of the present invention.

[0052] Figure 3 This is a schematic diagram of a flexible interconnection of a single-ring network in the backbone layer provided in Embodiment 2 of the present invention;

[0053] Figure 4 This is the original load curve diagram provided in Embodiment 2 of the present invention;

[0054] Figure 5 This is a diagram showing the planning and optimization results of the backbone layer single-ring network flexible interconnection provided in Embodiment 2 of the present invention;

[0055] Figure 6 This is a structural block diagram of a flexible interconnected backbone network planning system for medium-voltage distribution networks provided in Embodiment 3 of the present invention;

[0056] Figure 7 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention. Detailed Implementation

[0057] This invention provides a planning method, system, equipment, medium, and product for a flexible interconnected backbone network of medium-voltage distribution networks. It addresses the technical problem that existing planning methods for flexible interconnected backbone networks of medium-voltage distribution networks often employ a hierarchical optimization framework, decoupling medium- and long-term planning from short-term operation control. This results in a mismatch between the planning scheme and actual operating conditions, leading to low equipment utilization, poor investment efficiency, and reduced overall operational efficiency of the distribution network.

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1 , Figure 1 This is a flowchart illustrating the steps of a flexible interconnected backbone network planning method for medium-voltage distribution networks, as provided in Embodiment 1 of the present invention.

[0060] This invention provides a method for planning a flexible interconnected backbone network for medium-voltage distribution networks, comprising:

[0061] Step 101: Obtain the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and perform initial screening based on the candidate decision set to obtain multiple candidate schemes.

[0062] The candidate decision set refers to the set of candidate locations, capacity levels, and device types for flexible interconnection devices in a medium-voltage distribution network. The expression for the candidate location set is:

[0063] ;

[0064] in, For the set of candidate positions, The x-coordinate of the first position. The ordinate of the first position. The x-coordinate of the second position. The ordinate of the second position. For the first The x-coordinate of the position, For the first The y-coordinate of the position, This represents the total number of positions.

[0065] The selection of locations from the candidate set needs to consider multiple factors: First, priority should be given to locations of existing normally open tie switches, as these locations typically have good civil engineering conditions and grid operation experience; second, weak electrical links should be considered, such as long-term heavily loaded line sections and areas with low voltage; third, geographical conditions should be considered, such as corridors and the geographical location of substations. In actual engineering, The number is usually kept between 5 and 10 to ensure a reasonable range of enumerations.

[0066] The set of device types includes, but is not limited to, device types such as voltage source converters (VSC), flexible switches (FS), and solid-state transformers (SST). The specific expression for the set of device types is as follows:

[0067] ;

[0068] in, For device types It is the first type of device. For the second type of device, For the Nth t Device type.

[0069] Different device types have different technical characteristics and applicable scenarios. For example, the VSC type is suitable for scenarios requiring bidirectional high-power transmission and has reactive power compensation capabilities; the FS type is suitable for scenarios requiring only power regulation and has a lower cost; the SST type is suitable for scenarios requiring voltage level transformation. In actual planning, N t There are usually 2-3 types.

[0070] Electrical quantity data refers to a set of data that reflects the electrical characteristics and operating status of the power grid, including load correlation coefficients and peak-valley time offsets.

[0071] Candidate solutions refer to feasible configuration schemes that have passed both technical and economic screening and meet both operational constraints and investment budgets.

[0072] In this embodiment of the invention, a candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned are obtained. Based on preset device quantity constraints and capacity constraints, all combinations within the candidate decision set are traversed using an enumeration method to generate feasible solutions that meet the technical conditions. Each feasible solution is substituted into a preset investment cost function to calculate the corresponding investment cost, and feasible solutions whose investment costs do not exceed a preset budget threshold are selected as the corresponding candidate solutions.

[0073] Step 102: Generate the corresponding initial operation strategy population for each candidate scheme, and use the pre-trained prediction network and electrical quantity data to iteratively optimize the initial operation strategy population for each candidate scheme to obtain multiple target operation strategies and multiple target prediction data.

[0074] The initial operating strategy population refers to a set of operating strategies randomly generated according to operating constraints for a single candidate scheme. Each individual scheme corresponds to a combination of power regulation (i.e., power transmission in each time period), reactive power, and switching state.

[0075] Predictive networks refer to models trained on backpropagation neural networks to quickly predict power distribution network operation indicators, which can replace traditional power flow calculations and improve efficiency.

[0076] The target operation strategy refers to the optimal operation control strategy that is adapted to the corresponding candidate scheme after iterative optimization.

[0077] Target prediction data refers to key operational indicators such as the daily average load balance, daily equipment utilization, mutual balance strength, network loss reduction, and expected power shortage of the optimized system output by the prediction network in response to the target operation strategy.

[0078] In this embodiment of the invention, based on a genetic algorithm, initial operation decision populations corresponding to each candidate scheme are generated, using the transmission power, reactive power, and switch status of each time period as individuals. Corresponding regional load curves are generated for each individual in each initial operation strategy population. The regional load curves, candidate schemes, electrical quantity data, and switch statuses in the initial operation strategies are input into a pre-trained prediction network to obtain multiple prediction data. Each prediction data is input into a preset fitness function to obtain multiple fitness values. Each initial operation strategy population is iteratively optimized through selection, crossover, and mutation based on the fitness values ​​corresponding to each initial operation decision population. Finally, a set of optimal transmission power, reactive power, and switch statuses for each time period is output as the target operation strategy for each initial operation strategy population. Simultaneously, the prediction data corresponding to the target operation strategy is used as the target prediction data.

[0079] Step 103: Based on the prediction data of each target, screen each candidate scheme and each target operation strategy to obtain the corresponding flexible interconnection backbone network planning scheme.

[0080] The flexible interconnected backbone network planning scheme refers to the final planning result that includes the optimal device installation location, capacity, type, and supporting operation strategy.

[0081] In this embodiment of the invention, the network loss data corresponding to each target operation strategy is input into a preset network loss reduction benefit function to obtain multiple network loss reduction benefits. The expected power shortage of the preset baseline system is processed by differentiating it from the expected power shortage of the optimized system in each target prediction data to obtain multiple first differences. Each first difference is multiplied by the corresponding unit outage loss electricity price to obtain multiple outage loss reduction benefits. Investment data for each candidate scheme is obtained, and each investment data is input into a preset annual value function to obtain multiple investment cost penalty values. The target prediction data, network loss reduction benefits, outage loss reduction benefits, investment cost penalty values, delayed grid investment returns, and carbon reduction returns corresponding to each candidate scheme are input into a preset comprehensive benefit evaluation function to obtain multiple target benefit evaluation values. The candidate scheme and target operation strategy corresponding to the maximum value among the target benefit evaluation values ​​are selected as the corresponding flexible interconnection backbone network planning scheme.

[0082] It should be noted that the comprehensive benefit evaluation function is as follows:

[0083] ;

[0084] in, The target benefit assessment value, As the first evaluation weight coefficient, To reduce network losses and reduce efficiency, To reduce the benefits of power outage losses, To delay the return on power grid investment, As the penalty value for investment costs, This is the second evaluation weighting coefficient. The first technical weighting coefficient, To determine the daily average load balance, This is the second technical weighting coefficient. For equipment daily utilization rate, This is the third technical weighting coefficient. For mutual balance strength, This is the third evaluation weighting coefficient. This represents the carbon reduction benefit value.

[0085] In this embodiment of the invention, by acquiring the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, multiple candidate schemes are obtained through enumeration and initial screening based on the candidate decision set. A corresponding initial operating strategy population is generated for each candidate scheme, and a pre-trained prediction network and electrical quantity data are used to iteratively optimize the initial operating strategy populations corresponding to each candidate scheme, resulting in multiple target operating strategies and multiple target prediction data. Based on the target prediction data, each candidate scheme and each target operating strategy is screened to obtain the corresponding flexible interconnected backbone network planning scheme. This overcomes the technical problem that existing medium-voltage distribution network flexible interconnected backbone network planning methods often adopt a hierarchical optimization framework, decoupling medium- and long-term planning from short-term operation control, leading to a mismatch between the planned scheme and actual operating conditions, resulting in low equipment utilization, poor investment efficiency, and reduced overall operating efficiency of the distribution network. Compared with traditional planning methods, this approach quickly eliminates invalid schemes that do not meet the constraints through enumeration and initial screening. Then, iterative optimization of the operation strategy is carried out by relying on genetic algorithms, pre-trained prediction networks, and electrical quantity data. This achieves collaborative solution of network planning and operation strategy, ensuring the feasibility of candidate schemes in terms of quantity, capacity, and investment. At the same time, the optimization efficiency and accuracy of the operation strategy are greatly improved by using a high-precision prediction network. Furthermore, the screening of each candidate scheme and each target operation strategy based on the prediction data of each target ensures that the flexible interconnected backbone network planning scheme has technical rationality, economic advantages, and operational reliability. This effectively avoids the problem of planning and operation being disconnected and improves the overall operation efficiency of the distribution network.

[0086] Please see Figure 2 , Figure 2 This is a flowchart illustrating the steps of a flexible interconnection backbone network planning method for medium-voltage distribution networks provided in Embodiment 2 of the present invention.

[0087] This invention provides a method for planning a flexible interconnected backbone network for medium-voltage distribution networks, comprising:

[0088] Step 201: Obtain the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and perform initial screening based on the candidate decision set to obtain multiple candidate schemes.

[0089] Further, step 201 includes the following sub-steps:

[0090] S11. Based on the preset quantity constraints and capacity constraints, the feasible solutions corresponding to the candidate decision set are generated by enumeration.

[0091] The quantity constraint refers to the maximum number of flexible interconnect devices allowed to be installed under a planning scenario, which is determined jointly by the investment budget, power grid structure, and operation and management requirements. It is usually less than or equal to 3.

[0092] Capacity constraints refer to the minimum and maximum allowable capacity range for a single flexible interconnect unit, which must match regional load differences, line transmission capacity, and short-circuit capacity limitations. For example, common capacity constraints for 10kV systems are {1MW, 2MW, 3MW, 5MW, 8MW}; for 20kV systems, larger capacities may be required, such as {3MW, 5MW, 8MW, 10MW}. The range of capacity constraints should cover the minimum demand (typically 30% of the maximum regional load difference) and the maximum technically feasible value (limited by short-circuit capacity and equipment manufacturing capabilities).

[0093] In this embodiment of the invention, based on preset quantity and capacity constraints, an enumeration method is used to traverse all combinations of location, capacity, and device type in the candidate decision set, verifying each combination against the quantity and capacity constraints. Combinations that pass the verification are determined as feasible solutions corresponding to the candidate decision set. For example, when the quantity constraint is k, the set of feasible solutions can be represented as:

[0094] ;

[0095] in, This is a set of feasible solutions. For the selected position, combine vectors. For the selected capacity vector, For the selected type vector, Let k be the set of capacity levels, where k is the number of units allowed to be installed.

[0096] The expression for the total number of feasible solutions is as follows:

[0097] ;

[0098] in, Let K be the total number of feasible solutions, and K be the quantity constraint. The total number of positions, To allow for the total capacity when k units are installed, The total number of types allowed when k is installed. For example, when =5 (5 candidate positions), N c There are 4 (4 capacity levels), N t When there are 2 (2 types of devices) and the quantity constraint is 2:

[0099] When k=1: One option;

[0100] When k=2: One option;

[0101] Total number of feasible solutions: =40+640=680.

[0102] It should be noted that the enumeration method refers to a search method that performs a full combinatorial traversal of discrete variables such as position, capacity, and type in the candidate decision set to generate all possible configuration combinations.

[0103] S12. Input each feasible solution into the preset investment cost function to obtain multiple investment costs.

[0104] Investment cost refers to the total cost of the entire process, including equipment procurement, installation, civil engineering renovation, commissioning, and operation and maintenance, for a feasible solution.

[0105] In this embodiment of the invention, each feasible solution is substituted into a preset investment cost function to obtain multiple investment costs.

[0106] It should be noted that the investment cost function is as follows:

[0107] ;

[0108] in, Let i be the investment cost of the i-th device. For investment costs, For type correlation coefficient, For the capacity class of the i-th device, is the scale economy coefficient (usually 0.8-0.9), and i is the unit index.

[0109] S13. When the investment cost is less than or equal to the preset budget threshold, the feasible solution corresponding to the investment cost is determined as the candidate solution.

[0110] The budget threshold refers to the maximum affordable total investment set in advance for a planned project, and it serves as a criterion for judging the economic viability of a scheme.

[0111] In this embodiment of the invention, the investment cost corresponding to each feasible solution is compared one by one with a pre-set budget threshold. When the investment cost is greater than the budget threshold, the feasible solution corresponding to that investment cost is directly eliminated. When the investment cost is less than or equal to the pre-set budget threshold, it indicates that the solution is economically feasible, and the feasible solution corresponding to that investment cost is determined as a candidate solution that meets the requirements.

[0112] It should be noted that by generating feasible solutions through enumeration in the early stages of planning, combining quantity and capacity constraints, unreasonable decision combinations that do not meet equipment configuration limitations can be eliminated from the outset, ensuring that the solutions meet the basic engineering implementation conditions. Then, each feasible solution is quantitatively calculated using an investment cost function, and economic screening is performed based on budget thresholds. This achieves a dual verification of technical feasibility and economic rationality, preventing invalid solutions from entering subsequent complex optimization processes to reduce computational overhead, and ensuring that the final selected candidate solutions all meet actual engineering constraints. This lays a compliant and efficient foundation for the subsequent collaborative optimization of grid planning and operation strategies.

[0113] Step 202: Generate the corresponding initial operating strategy population for each candidate scheme, and generate the corresponding regional load curve for each individual in each initial operating strategy population, wherein each individual corresponds to an initial operating strategy.

[0114] Regional load curves refer to the load power curves of each power supply area of ​​a distribution network over time under a certain operating strategy.

[0115] The initial operating strategy refers to the original operating control scheme that has not been optimized, including the transmission power, reactive power and switching status at each time period.

[0116] In this embodiment of the invention, for each candidate scheme, an initial operating strategy population containing multiple individual operating strategies is randomly generated, constrained by the distribution network safety operation boundary, the control range of flexible interconnection devices, and the switching operation rules. Based on the transmission power and reactive power of each individual in each initial operating strategy population at different time periods, the regional load curve corresponding to that individual under the initial operating strategy is simulated and calculated.

[0117] Step 203: Input the regional load curve, candidate scheme, electrical quantity data and switch status in the initial operation strategy corresponding to each individual into the pre-trained prediction network to obtain multiple prediction data.

[0118] In this embodiment of the invention, the regional load curve, candidate scheme, electrical quantity data and switch status in the initial operation strategy corresponding to each individual are respectively input into a pre-trained prediction network for forward calculation and index inference to obtain multiple prediction data.

[0119] It should be noted that the training process of the prediction network is as follows:

[0120] A1. Obtain multiple training network planning data, preprocess each training network planning data, and obtain the corresponding prediction feature set.

[0121] Training network planning data refers to sample data used to train the prediction network, which includes planning and operation-related information such as device location, capacity, type, regional load curve, switch status, electrical quantity data, and operating indicators.

[0122] The prediction feature set refers to the set of feature vectors extracted after preprocessing that can be directly input into a neural network; it is the training input data for the prediction network.

[0123] Preprocessing refers to standardized processing operations such as cleaning, normalizing, denoising, and unifying dimensions of the original training data.

[0124] In this embodiment of the invention, multiple sets of training network planning data are collected through historical engineering cases, simulation examples and typical operating scenarios. Then, each set of training network planning data is normalized, outlier removal, dimension alignment and feature extraction are performed to obtain the corresponding prediction feature set.

[0125] A2. Input the predicted feature set into the preset error backpropagation neural network for training to obtain the corresponding training prediction data.

[0126] Backpropagation neural networks (BP neural networks) are classic supervised learning networks that achieve fitting predictions by forward propagating the output and backpropagating to correct the weights. The specific structural design of a BP neural network is as follows:

[0127] Input layer: 210 neurons (192-dimensional payload data + 18-dimensional feature parameters)

[0128] Hidden layer 1: 256 neurons, ReLU activation function, Dropout rate 0.3

[0129] Hidden layer 2: 128 neurons, ReLU activation function, Dropout rate 0.3

[0130] Hidden layer 3: 64 neurons, activation function tanh

[0131] Output layer: 7 neurons (3 technical indicators + 4 operational performance indicators).

[0132] Training prediction data refers to the prediction index data output by the BP neural network during the training process, which is used to compare with the true value to calculate the error.

[0133] In this embodiment of the invention, the predicted feature set is input into a preset error backpropagation neural network for training, and the output result is calculated through forward propagation of the error backpropagation neural network to obtain training prediction data corresponding to the predicted feature set.

[0134] A3. Based on the preset loss function, determine the loss function value of the prediction feature set according to the training prediction data.

[0135] In this embodiment of the invention, the training prediction data and the prediction feature set are input into a preset loss function to obtain the corresponding loss function value.

[0136] It should be noted that the loss function is as follows:

[0137] ;

[0138] in, The value of the loss function. This represents the true daily average load balance of the k1-th sample. The daily average load balance of the k1th sample. This represents the true daily utilization rate of the equipment for the k1th sample. Let k1 be the daily utilization rate of the equipment. This represents the true value of the mutual balance strength for the k1th sample. Let k1 be the mutual balance strength of the sample. Let the annual operating costs be the true values ​​for k1 samples. Let N be the annual operating cost for k1 samples, where k1 is the sample index and N is the number of samples. b The total number of samples.

[0139] A4. When the loss function value is greater than or equal to the preset loss threshold, adjust the network parameters of the error backpropagation neural network until the loss function value is less than the loss threshold.

[0140] The loss threshold refers to a pre-set upper limit of error, which is the criterion for determining whether neural network training has reached the target and whether training should be terminated.

[0141] Network parameters refer to the weight coefficients and bias terms in the error backpropagation neural network, and are the core parameters for the network to achieve the function of fitting and prediction.

[0142] In this embodiment of the invention, when the loss function value is greater than or equal to the preset loss threshold, it indicates that the current network prediction accuracy does not meet the requirements. Then, the weights and bias parameters of the neural network are adjusted layer by layer through the error backpropagation mechanism, and the prediction feature set is re-inputted for training and loss function value calculation. The parameter update and error verification process is repeatedly executed until the loss function value is less than the loss threshold, so that the network prediction accuracy meets the planning requirements.

[0143] A5. When the loss function value is less than the loss threshold, the corresponding prediction network is generated.

[0144] In this embodiment of the invention, when the loss function value is less than the loss threshold, it indicates that the network has completed sufficient learning and the prediction accuracy meets the planning requirements. At this time, parameter adjustment and iterative training are stopped, and the error of the currently trained neural network is backpropagated to determine the prediction network.

[0145] Step 204: Input each predicted data into the preset fitness function to obtain multiple fitness values.

[0146] Fitness refers to the evaluation value output by the fitness function. The higher the value, the better the overall effect of the corresponding running strategy.

[0147] In this embodiment of the invention, each predicted data is substituted into a preset fitness function to obtain multiple fitness values.

[0148] It should be noted that the fitness function is as follows:

[0149] ;

[0150] in, For fitness, To predict comprehensive indicators, Let be the voltage of the i1th node at time t. The first penalty coefficient, This is the second penalty coefficient. Let i be the load rate of the line from i1 to j at time t, where i1 is the first index of the node and j is the second index of the node.

[0151] Step 205: Optimize each initial running strategy population according to each fitness level to obtain multiple target running strategies.

[0152] In this embodiment of the invention, fitness is used as the evaluation criterion. According to the principle that the higher the fitness, the better the individual, selection, crossover and mutation operations are performed on each initial running strategy population in sequence. Through multiple rounds of iteration, inferior strategies are continuously eliminated and superior strategies are retained and evolved, so that the population continues to converge towards the optimal running direction. Finally, the optimal target running strategy is generated and output for each candidate scheme.

[0153] It should be noted that selection refers to the evolutionary operation of selecting high-fitness individuals from the population to enter the next generation, implemented through tournament selection (tournament size=3). Crossover refers to the evolutionary operation of combining parameters of superior individuals to generate new strategy individuals, implemented through simulated binary crossover (SBX) with a crossover probability of 0.8. Mutation refers to the operation of making small random adjustments to individual parameters to improve population diversity, implemented through Gaussian mutation with an initial mutation probability of 0.1, which adaptively adjusts with each generation.

[0154] It should be noted that by generating corresponding regional load curves for each individual in the initial operating strategy population, the abstract operating strategy is transformed into quantifiable load data, providing an input basis that fits the actual operating scenario for subsequent predictions. Secondly, the regional load curves, candidate schemes, electrical quantity data, and switch status data are jointly input into a pre-trained prediction network to ensure that the prediction data comprehensively reflects the operating status of the schemes and improves prediction accuracy. Then, the prediction data is quantitatively evaluated using a fitness function, providing a clear basis for judging the merits of the operating strategy population optimization. Subsequently, the population is iteratively optimized based on fitness, gradually eliminating inferior strategies and evolving superior strategies, ultimately obtaining the target operating strategy that fits each candidate scheme. Simultaneously, the prediction data corresponding to the target operating strategy is determined as the target prediction data. This ensures the optimality of the target operating strategy and provides high-precision, high-reference-value quantitative indicators for subsequent scheme selection, effectively improving the efficiency and rationality of operating strategy optimization, avoiding blind optimization, and achieving coordinated adaptation between the operating strategy and the network planning, providing solid support for the final selection of flexible interconnected backbone network planning schemes.

[0155] Step 206: Use the prediction data corresponding to each target operation strategy as the target prediction data.

[0156] In this embodiment of the invention, the prediction data corresponding to each target operation strategy is determined as the target prediction data.

[0157] Step 207: Based on the prediction data of each target, screen each candidate scheme and each target operation strategy to obtain the corresponding flexible interconnection backbone network planning scheme.

[0158] Furthermore, step 207 includes the following sub-steps:

[0159] S21. Based on the preset technical weights, perform weighted calculations on the prediction data of each target to obtain multiple technical benefit values.

[0160] Technical weight refers to a coefficient set in advance according to planning requirements to characterize the importance of each technical indicator.

[0161] Technical benefit value refers to a quantitative evaluation value obtained by integrating multiple operational indicators and used to measure the technical merits of a solution.

[0162] In this embodiment of the invention, based on preset technical weights, the daily average load balance, daily equipment utilization rate and mutual balance intensity in each target prediction data are weighted and calculated to obtain multiple technical benefit values.

[0163] S22. Input the network loss data corresponding to each target operation strategy into the preset network loss reduction benefit function to obtain multiple network loss reduction benefits.

[0164] Network loss reduction benefits refer to the direct economic benefits generated by the reduction of network loss after configuring flexible interconnection devices.

[0165] Network loss data refers to the active network loss at various times after the flexible interconnection device is configured.

[0166] In this embodiment of the invention, the network loss data corresponding to each target operation strategy is input into a preset network loss reduction benefit function, and the network loss difference between the flexible interconnection scheme and the original network structure is quantitatively compared to obtain multiple network loss reduction benefits.

[0167] It should be noted that the benefit function for reducing network loss is as follows:

[0168] ;

[0169] in, To reduce network losses and reduce efficiency, Let t be the active power loss before optimization. Let be the active power loss at time t. The unit price of electricity. Let be the weighting coefficient at time t.

[0170] S23. The expected power shortage of the preset baseline system and the expected power shortage of the optimized system in each target prediction data are processed by difference to obtain multiple first differences.

[0171] The baseline system expected power shortage refers to the original expected power shortage of the distribution network within the planning period when flexible interconnection devices are not installed.

[0172] Optimizing the expected power shortage of the system refers to the reduction in the expected power shortage of the system after adopting the corresponding candidate schemes and target operating strategies.

[0173] In this embodiment of the invention, the difference between the preset baseline system expected power shortage and the optimized system expected power shortage in each target prediction data is calculated to obtain multiple first differences.

[0174] S24. Multiply each of the first differences by the corresponding unit power outage loss electricity price to obtain multiple power outage loss reduction benefits.

[0175] Unit power outage loss price refers to the price charged for the economic loss to users caused by a power outage per unit of electricity, and is used to convert the power shortage into economic benefits.

[0176] The benefits of reduced power outage losses refer to the direct economic gains generated by flexible interconnection, which improves power supply reliability and reduces power outages.

[0177] In this embodiment of the invention, the multiplication between each first difference and the corresponding unit power outage loss electricity price is calculated to obtain multiple power outage loss reduction benefits.

[0178] In another embodiment, the preset baseline system expected power shortage and the optimized system expected power shortage in each target prediction data are respectively input into a preset power outage loss reduction benefit function to obtain multiple power outage loss reduction benefits.

[0179] It should be noted that the benefit function for reducing power outage losses is as follows:

[0180] ;

[0181] in, To reduce the benefits of power outage losses, The expected power shortage for the baseline system. To optimize the system's expected power supply shortage, The unit is the electricity price for power outage losses.

[0182] S25. The benefits of each network loss reduction are summed with the corresponding benefits of power outage loss reduction and delayed power grid investment returns to obtain multiple economic benefit values.

[0183] Delaying the return on power grid investment refers to the fact that by configuring flexible interconnection devices (FIDs) in medium-voltage distribution networks, the rigid investments that would otherwise be required due to load growth, such as line expansion, transformer capacity increase, and new / upgraded substations, can be postponed or even avoided through feeder power sharing and load peak shaving and valley filling. The resulting savings in investment costs and the time value of money are converted into annualized economic benefits.

[0184] In this embodiment of the invention, the sum of each network loss reduction benefit and the corresponding power outage loss reduction benefit and delayed power grid investment return is calculated to obtain multiple economic benefit values.

[0185] S26. Obtain the investment data of each candidate scheme, and input each investment data into the preset equal annual value function to obtain multiple investment cost penalty values.

[0186] Investment data refers to all relevant data related to the purchase, installation, commissioning, operation and maintenance of equipment for the candidate solutions, including but not limited to total investment cost, annual operating cost, discount rate and project life.

[0187] In this embodiment of the invention, investment data for each candidate scheme is obtained, and each investment data is input into a preset equal-annual value function to convert the one-time total investment into an average annual cost form, thereby obtaining multiple investment cost penalty values.

[0188] It should be noted that the isochronous value function is specifically as follows:

[0189] ;

[0190] in, As the penalty value for investment costs, The total investment cost, This is the capital recovery coefficient. Annual operating costs, Let n be the discount rate and n be the project lifespan.

[0191] S27. Based on the preset evaluation weights, the technical benefit value, economic benefit value, investment cost penalty value and carbon reduction benefit value corresponding to each candidate scheme are weighted and calculated to obtain multiple target benefit evaluation values.

[0192] Carbon reduction benefit value refers to the annual benefit derived from the reduction of carbon emissions due to reduced grid losses, increased absorption of new energy sources, and delayed expansion of fossil fuel power generation / grid capacity after configuring a flexible interconnected backbone network, which is converted into the economic value of carbon emission reduction.

[0193] In this embodiment of the invention, based on preset evaluation weights, the technical benefit value, economic benefit value, investment cost penalty value, and carbon reduction benefit value corresponding to each candidate scheme are weighted and calculated to obtain multiple target benefit evaluation values. For example, the target benefit evaluation value = first evaluation weight coefficient * economic benefit value - investment cost penalty value + second evaluation weight coefficient * technical benefit value + third technical weight coefficient * carbon reduction benefit value.

[0194] S28. Select the candidate scheme and target operation strategy corresponding to the maximum value among the various target benefit evaluation values ​​as the corresponding flexible interconnection backbone network planning scheme.

[0195] In this embodiment of the invention, by comparing the target benefit evaluation values ​​of all candidate schemes, the candidate scheme with the largest evaluation value and the matching target operation strategy are selected as the final flexible interconnection backbone network planning scheme.

[0196] It should be noted that by quantifying the benefits of technology, network loss reduction, power outage loss reduction, and carbon reduction in stages, the abstract advantages of the proposed solutions are transformed into quantifiable indicators that can be accurately compared. This avoids the bias of subjective judgment and makes the comparison of the merits of each candidate solution clearer and more objective, ensuring the fairness of the evaluation results. At the same time, by using preset evaluation weights as a link, multi-dimensional indicators of technology, economy, cost, and environmental protection (carbon reduction) are integrated and weighted, breaking the limitations of single-dimensional evaluation. This ensures that the final selected solution not only meets the requirements of technical feasibility but also possesses economic rationality and environmental compliance, truly achieving the planning goal of "technologically feasible, economically optimal, cost-controllable, and green and low-carbon." Fourth, the entire evaluation process is logically coherent and the steps are clear. From single benefit calculation to comprehensive weighting, from cost penalty to final selection, a closed-loop evaluation system is formed. This not only improves the efficiency and accuracy of scheme selection, but also ensures that the final flexible interconnected backbone network planning scheme can adapt to the actual operation needs of the medium-voltage distribution network, taking into account both practicality and long-term effectiveness. It effectively avoids the problems of disconnect between planning and actual operation and imbalance between benefits and costs, and provides scientific support for the flexible and intelligent development of the medium-voltage distribution network.

[0197] For example, Case 1 considers a typical urban 10kV single-ring network. Referring to Table 1 (Basic Parameters of Case 1 System) and Table 2 (Load Timing Characteristics of Case 1, Partially Shown), the system consists of two feeders (F1 and F2), each connected to a 10kV / 0.4kV distribution transformer. The feeders are connected by a normally open tie switch. A schematic diagram of the flexible interconnection of the main layer single-ring network is shown below. Figure 3 As shown, the two feeders serve the commercial area and the residential area respectively, and their load characteristics are clearly complementary. The original load curve is as follows. Figure 4 As shown.

[0198] Table 1

[0199]

[0200] Table 2

[0201]

[0202] The two feeder lines exhibit distinctly complementary load characteristics: F1 (commercial area) experiences higher load during working hours (8:00-18:00), while F2 (residential area) experiences higher load in the evening (18:00-22:00). The load correlation coefficient is -0.62, indicating good mutual support potential. Furthermore, the optimization results obtained in this invention are as follows: Figure 5 As shown, this invention is superior to the prior art, wherein, Figure 5The diagram includes four sub-graphs (a) to (d), visually presenting the effectiveness and optimization results of the planning method from four dimensions: scheme selection, algorithm convergence, load improvement, and power balance. Figure (a) is a ranking of the comprehensive benefits of candidate schemes, comparing the technical benefits, economic benefits, carbon reduction benefits, and investment cost of each enumerated scheme. The results show that the optimal scheme significantly outperforms other candidate schemes, improving upon the second-best scheme by 12%–18%, achieving accurate selection of the globally optimal scheme. Figure (b) is the iterative convergence curve of the genetic algorithm, with the number of iterations on the horizontal axis and the fitness of the running strategy on the vertical axis, comparing the changes in strategy quality during optimization. The fitness rapidly converged from an initial 0.42 to 0.8139, an improvement of 93.8%, and stabilized after 20 generations, verifying the efficiency and stability of the running strategy optimization algorithm. Figure (c) is a comparison of the regional load curves before and after optimization. Focusing on the two core dimensions of load balance and peak-valley difference, the optimized load balance improved from 0.58 to 0.91 (an increase of 56.9%), the peak-valley difference decreased from 4.2MW to 1.8MW (a decrease of 57.1%), and the average daily network loss decreased by 9.7%, effectively improving the problem of uneven load distribution. Figure (d) shows the power exchange curve of the flexible interconnection device, comparing the power mutual assistance capability, equipment utilization rate, and mutual assistance balance strength. After optimization, the mutual assistance balance strength improved from 0.61 to 0.93 (an increase of 52.5%), the daily equipment utilization rate improved from 58% to 89% (an increase of 53.4%), the bidirectional power mutual assistance was smooth and stable, with no risk of overload or reverse power flow, and the power mutual assistance efficiency of the flexible interconnection device was fully utilized.

[0203] In this embodiment of the invention, by acquiring the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, multiple candidate schemes are obtained through enumeration and initial screening based on the candidate decision set. A corresponding initial operating strategy population is generated for each candidate scheme, and a pre-trained prediction network and electrical quantity data are used to iteratively optimize the initial operating strategy populations corresponding to each candidate scheme, resulting in multiple target operating strategies and multiple target prediction data. Based on the target prediction data, each candidate scheme and each target operating strategy is screened to obtain the corresponding flexible interconnected backbone network planning scheme. This overcomes the technical problem that existing medium-voltage distribution network flexible interconnected backbone network planning methods often adopt a hierarchical optimization framework, decoupling medium- and long-term planning from short-term operation control, leading to a mismatch between the planned scheme and actual operating conditions, resulting in low equipment utilization, poor investment efficiency, and reduced overall operating efficiency of the distribution network. Compared with traditional planning methods, this approach quickly eliminates invalid schemes that do not meet the constraints through enumeration and initial screening. Then, iterative optimization of the operation strategy is carried out by relying on genetic algorithms, pre-trained prediction networks, and electrical quantity data. This achieves collaborative solution of network planning and operation strategy, ensuring the feasibility of candidate schemes in terms of quantity, capacity, and investment. At the same time, the optimization efficiency and accuracy of the operation strategy are greatly improved by using a high-precision prediction network. Furthermore, the screening of each candidate scheme and each target operation strategy based on the prediction data of each target ensures that the flexible interconnected backbone network planning scheme has technical rationality, economic advantages, and operational reliability. This effectively avoids the problem of planning and operation being disconnected and improves the overall operation efficiency of the distribution network.

[0204] Please see Figure 6 , Figure 6 This is a structural block diagram of a flexible interconnected backbone network planning system for medium-voltage distribution networks provided in Embodiment 3 of the present invention.

[0205] This invention provides a flexible interconnection backbone network planning system for medium-voltage distribution networks, comprising:

[0206] The acquisition module 301 is used to acquire the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and to perform enumeration and preliminary screening based on the candidate decision set to obtain multiple candidate schemes;

[0207] The prediction module 302 is used to generate a corresponding initial operation strategy population based on each candidate scheme, and to iteratively optimize the initial operation strategy population corresponding to each candidate scheme using a pre-trained prediction network and electrical quantity data to obtain multiple target operation strategies and multiple target prediction data.

[0208] The filtering module 303 is used to filter each candidate scheme and each target operation strategy based on the prediction data of each target to obtain the corresponding flexible interconnection backbone network planning scheme.

[0209] Furthermore, the acquisition module 301 includes:

[0210] The enumeration submodule is used to generate feasible solutions corresponding to candidate decision sets based on preset quantity constraints and capacity constraints using an enumeration method.

[0211] The investment submodule is used to input each feasible solution into a preset investment cost function to obtain multiple investment costs;

[0212] The first screening submodule is used to identify feasible solutions corresponding to investment costs as candidate solutions when the investment cost is less than or equal to a preset budget threshold.

[0213] Furthermore, the prediction module 302 includes:

[0214] A submodule is constructed to generate corresponding regional load curves for each individual in the population of each initial running strategy, where each individual corresponds to an initial running strategy.

[0215] The prediction submodule is used to input the regional load curve, candidate scheme, electrical quantity data and switch status in the initial operation strategy corresponding to each individual into the pre-trained prediction network to obtain multiple prediction data.

[0216] The adaptation submodule is used to input each prediction data into a preset fitness function to obtain multiple fitness values;

[0217] The optimization submodule is used to optimize each initial running strategy population according to each fitness level to obtain multiple target running strategies;

[0218] The predicted data corresponding to each target's operational strategy is used as the target prediction data.

[0219] Furthermore, the filtering module 303 includes:

[0220] The technology benefit submodule is used to perform weighted calculations on the prediction data of each target based on preset technology weights to obtain multiple technology benefit values.

[0221] The network loss reduction benefit submodule is used to input the network loss data corresponding to each target operation strategy into a preset network loss reduction benefit function to obtain multiple network loss reduction benefits;

[0222] The power outage loss reduction benefit submodule is used to perform difference processing between the preset baseline system expected power shortage and the optimized system expected power shortage in each target prediction data to obtain multiple first differences;

[0223] Each first difference is multiplied by the corresponding unit power outage loss electricity price to obtain multiple power outage loss reduction benefits;

[0224] The economic benefit value submodule is used to sum up each network loss reduction benefit with the corresponding power outage loss reduction benefit and delayed grid investment returns to obtain multiple economic benefit values.

[0225] The investment cost penalty value submodule is used to obtain investment data for each candidate scheme, and input each investment data into a preset equal annual value function to obtain multiple investment cost penalty values;

[0226] The target benefit assessment submodule is used to perform weighted calculations on the technical benefit value, economic benefit value, investment cost penalty value and carbon reduction benefit value of each candidate scheme based on preset assessment weights, so as to obtain multiple target benefit assessment values.

[0227] The second screening submodule is used to select the candidate scheme and target operation strategy corresponding to the maximum value among the various target benefit evaluation values ​​as the corresponding flexible interconnection backbone network planning scheme.

[0228] Furthermore, the training process of the prediction network is as follows:

[0229] Multiple training network planning data are obtained, and each training network planning data is preprocessed to obtain the corresponding prediction feature set;

[0230] The predicted feature set is input into a pre-defined error backpropagation neural network for training to obtain the corresponding training prediction data.

[0231] Based on a preset loss function, the loss function value of the prediction feature set is determined according to the training prediction data;

[0232] When the loss function value is greater than or equal to the preset loss threshold, the network parameters of the error backpropagation neural network are adjusted until the loss function value is less than the loss threshold.

[0233] When the loss function value is less than the loss threshold, the corresponding prediction network is generated.

[0234] Furthermore, the isochronous value function is specifically as follows:

[0235] ;

[0236] in, As the penalty value for investment costs, The total investment cost, This is the capital recovery coefficient. Annual operating costs, Let n be the discount rate and n be the project lifespan.

[0237] Please see Figure 7 , Figure 7This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention.

[0238] An electronic device according to an embodiment of the present invention includes: a memory 401 and a processor 402. The memory 401 stores a computer program. When the computer program is executed by the processor 402, the processor 402 executes the medium-voltage distribution network flexible interconnection backbone network planning method as described in any of the above embodiments.

[0239] Memory 401 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Memory 401 has storage space 403 for program code 413 for performing any of the method steps described above. For example, storage space 403 for program code may include individual program codes 413 for implementing the various steps in the methods described above. This program code may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When run by a computing processing device, this code causes the computing processing device to perform the various steps in the methods described above. This program code may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When these codes are run by a computing processing device, the device causes it to execute the various steps in the medium-voltage distribution network flexible interconnection backbone network planning method described above.

[0240] Embodiment 5 of the present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the medium-voltage distribution network flexible interconnection backbone network planning method as described in any of the above embodiments.

[0241] Embodiment 6 of the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer performs the medium-voltage distribution network flexible interconnection backbone network planning method as described in any of the above embodiments.

[0242] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0243] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0244] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0245] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0246] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0247] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A planning method for a flexible interconnected backbone network of a medium-voltage distribution network, characterized in that, include: Obtain the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and perform initial screening based on the candidate decision set to obtain multiple candidate schemes; Each candidate scheme generates a corresponding initial operating strategy population, and the pre-trained prediction network and the electrical quantity data are used to iteratively optimize the initial operating strategy population corresponding to each candidate scheme to obtain multiple target operating strategies and multiple target prediction data. Based on the predicted data for each target, the candidate schemes and the operational strategies for each target are screened to obtain the corresponding flexible interconnected backbone network planning scheme.

2. The planning method for flexible interconnected backbone network of medium-voltage distribution network according to claim 1, characterized in that, The step of enumerating and initially screening based on the candidate decision set to obtain multiple candidate solutions includes: Based on preset quantity and capacity constraints, feasible solutions corresponding to the candidate decision set are generated by enumeration. Each of the feasible solutions is input into a preset investment cost function to obtain multiple investment costs; When the investment cost is less than or equal to a preset budget threshold, the feasible solution corresponding to the investment cost is determined as a candidate solution.

3. The planning method for flexible interconnection backbone network of medium-voltage distribution network according to claim 1, characterized in that, The step of iteratively optimizing the initial operating strategy population corresponding to each candidate scheme using a pre-trained prediction network and the electrical quantity data to obtain multiple target operating strategies and multiple target prediction data includes: Each individual in the population of each initial operating strategy is given a corresponding regional load curve, wherein each individual corresponds to an initial operating strategy. The regional load curves, candidate schemes, electrical quantity data, and switch states in the initial operation strategy corresponding to each individual are respectively input into a pre-trained prediction network to obtain multiple prediction data. Each of the predicted data is input into a preset fitness function to obtain multiple fitness values; Each initial running strategy population is optimized based on its fitness to obtain multiple target running strategies; The predicted data corresponding to each of the target operation strategies is used as the target prediction data.

4. The method for planning a flexible interconnected backbone network for medium-voltage distribution networks according to claim 1, characterized in that, The step of filtering each candidate scheme and each target operation strategy based on each target prediction data to obtain the corresponding flexible interconnection backbone network planning scheme includes: Based on preset technical weights, the predicted data of each target are weighted and calculated to obtain multiple technical benefit values; The network loss data corresponding to each of the target operation strategies is input into a preset network loss reduction benefit function to obtain multiple network loss reduction benefits; The expected power shortage of the preset baseline system is compared with the expected power shortage of the optimized system in each of the target prediction data to obtain multiple first differences; Each of the first differences is multiplied by the corresponding unit power outage loss electricity price to obtain multiple power outage loss reduction benefits; The benefits of reducing network losses are summed with the corresponding benefits of reducing power outage losses and delaying the return on power grid investment to obtain multiple economic benefit values. The investment data of each candidate scheme is obtained, and each investment data is input into a preset equal annual value function to obtain multiple investment cost penalty values; Based on preset evaluation weights, the technical benefit value, economic benefit value, investment cost penalty value and carbon reduction benefit value corresponding to each candidate scheme are weighted and calculated to obtain multiple target benefit evaluation values; The candidate scheme and target operation strategy corresponding to the maximum value among the various target benefit evaluation values ​​are selected as the corresponding flexible interconnection backbone network planning scheme.

5. The planning method for a flexible interconnected backbone network of a medium-voltage distribution network according to claim 1, characterized in that, The training process of the prediction network is as follows: Multiple training network planning data are acquired, and each training network planning data is preprocessed to obtain the corresponding prediction feature set. The predicted feature set is input into a preset error backpropagation neural network for training to obtain the corresponding training prediction data; Based on a preset loss function, the loss function value of the prediction feature set is determined according to the training prediction data; When the loss function value is greater than or equal to a preset loss threshold, the network parameters of the backpropagation neural network are adjusted until the loss function value is less than the loss threshold. When the loss function value is less than the loss threshold, a corresponding prediction network is generated.

6. The method for planning a flexible interconnected backbone network for medium-voltage distribution networks according to claim 4, characterized in that, The isochronous value function is specifically as follows: ； in, As the penalty value for investment costs, The total investment cost, This is the capital recovery coefficient. Annual operating costs, Let n be the discount rate and n be the project lifespan.

7. A flexible interconnected backbone network planning system for medium-voltage distribution networks, characterized in that, include: The acquisition module is used to acquire the candidate decision set and electrical quantity data of the medium-voltage distribution network to be planned, and to perform enumeration and initial screening based on the candidate decision set to obtain multiple candidate schemes; The prediction module is used to generate a corresponding initial operation strategy population based on each of the candidate schemes, and to iteratively optimize the initial operation strategy populations corresponding to each of the candidate schemes using a pre-trained prediction network and the electrical quantity data, so as to obtain multiple target operation strategies and multiple target prediction data. The filtering module is used to filter each candidate scheme and each target operation strategy based on the prediction data of each target, so as to obtain the corresponding flexible interconnection backbone network planning scheme.

8. An electronic device, characterized in that, The system includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the medium-voltage distribution network flexible interconnection backbone network planning method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the medium-voltage distribution network flexible interconnection backbone network planning method as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, wherein when the program instructions are executed by a computer, the computer performs the medium-voltage distribution network flexible interconnection backbone network planning method as described in any one of claims 1-6.

Images