Methods, devices, and electronic equipment for obtaining power distribution network operation plans

Abstract

This application discloses a method, apparatus, and electronic device for obtaining distribution network operation schemes. Relating to the field of distribution network power dispatching, the method includes: acquiring multiple sets of historical operation data within a preset period of the distribution network; calculating a first per-unit value and a second per-unit value based on the historical operation data; converting a target model into model constraints, obtaining operation constraints, and combining the model constraints and operation constraints into a constraint set; solving the constraint set with the goal of minimizing the operating cost of the distribution network to obtain a distribution network operation scheme; generating multiple undetermined distribution network operation schemes from the distribution network operation schemes, determining the constraint satisfaction probability of the distribution network operation schemes, and determining the target distribution network operation scheme from the multiple undetermined distribution network operation schemes and the distribution network operation schemes based on the constraint satisfaction probability. This application solves the problem of low efficiency in obtaining real-time operation schemes of distribution networks in related technologies.

Description

Technical Field

[0001] This application relates to the field of power dispatching in power distribution networks, and more specifically, to a method, apparatus, and electronic equipment for obtaining power distribution network operation plans. Background Technology

[0002] In related technologies, as the topological complexity of urban power distribution networks gradually increases and the requirements for safe operation become more stringent, power flow calculations become increasingly complex. Urban power distribution network power flow calculations primarily rely on Kirchhoff's circuit laws to construct network voltage and current balance constraints. Then, iterative solutions, such as the Newton-Raphson algorithm, are used for iterative solving. This algorithm offers extremely high calculation accuracy but requires complete and accurate line impedance parameters.

[0003] With the large-scale integration of high-power loads such as electric vehicles into distribution networks, urban distribution networks require continuous upgrades and modifications to their lines and equipment. This leads to increasing complexity in urban distribution network topology and equipment, making it increasingly difficult to accurately obtain distribution network impedance parameters. On the one hand, the excessive number of distribution network nodes and frequency and voltage regulation equipment makes it increasingly difficult to solve power flow problems based on physical models. On the other hand, due to the difficulty in accurately obtaining distribution network impedance parameters, the calculated power flow may deviate significantly from reality, making it difficult to guarantee safe operation. This makes real-time optimization of distribution network operation increasingly challenging. However, the increasingly complex distribution network structure also urgently requires more precise optimization operation schemes to guide efficient and safe operation. Therefore, distribution network operation schemes in related technologies require detailed calculations of distribution network topology and impedance data, which are difficult to obtain in practical applications. Furthermore, the obtained distribution network topology data may deviate significantly from actual operating conditions, failing to fully meet the constraints of safe operation. Additionally, the algorithms for solving distribution network operation schemes in related technologies have excessively long solution times, making it difficult to meet the real-time requirements of distribution network optimization operation.

[0004] There is currently no effective solution to the problem of low efficiency in obtaining real-time operation plans for power distribution networks in related technologies. Summary of the Invention

[0005] The main objective of this application is to provide a method, apparatus, and electronic device for obtaining distribution network operation schemes, so as to solve the problem of low efficiency in obtaining real-time operation schemes of distribution networks in related technologies.

[0006] To achieve the above objectives, according to one aspect of this application, a method for obtaining a distribution network operation scheme is provided. The method includes: acquiring multiple sets of historical operation data within a preset period of the distribution network; calculating a first per-unit value and a second per-unit value based on each set of historical operation data, wherein the first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network; converting a target model into model constraints, and acquiring the operation constraints of the distribution network; combining the model constraints and the operation constraints into a constraint set, wherein the target model is trained from multiple sets of training samples, each... The training samples include a set of historical operating data and a set of first and second per-unit values ​​of the historical operating data. With the goal of minimizing the operating cost of the distribution network, the set of constraints is solved based on the optimization objective to obtain a distribution network operation scheme. This scheme includes at least the distribution network's energy storage capacity, power generation capacity, and power purchase capacity. Multiple undetermined distribution network operation schemes are generated from these schemes. The constraint satisfaction probability of each undetermined scheme is determined based on these schemes. Finally, the target distribution network operation scheme is determined from the multiple undetermined schemes and the established distribution network operation schemes based on the constraint satisfaction probability.

[0007] Optionally, calculating the first and second per-unit values ​​of historical operating data based on each set of historical operating data includes: determining the node voltage of each node in the distribution network and the line current of each line; determining the maximum and minimum node voltages in the voltage boundary constraints of the distribution network; and determining the maximum line current in the current boundary constraints of the distribution network; for each node, calculating the first difference between the maximum node voltage and the node voltage of the node, and calculating the ratio of the first difference to the rated voltage of the distribution network to obtain the first voltage per-unit value of the node; calculating the ratio of the node voltage of the node to the minimum node voltage of the node. The second voltage difference is calculated, and the ratio of the second voltage difference to the rated voltage of the distribution network is calculated to obtain the second voltage per-unit value of the node; the minimum voltage per-unit value is determined from the first voltage per-unit value and the second voltage per-unit value of all nodes, and the minimum voltage per-unit value is determined as the first per-unit value; for each line, the third difference between the maximum line current and the line current is calculated, and the ratio of the third difference to the maximum line current is calculated to obtain the current per-unit value of the line; the minimum current per-unit value is determined from the current per-unit values ​​of all lines, and the minimum current per-unit value is determined as the second per-unit value.

[0008] Optionally, converting the target model into model constraints includes: determining the input layer, hidden layer, and output layer of the target model, and determining the model parameters of the target model; determining the expression of the input layer as the first constraint, the linear mapping expression of the hidden layer as the second constraint, the output expression of the activation function of the hidden layer as the third constraint, and the expression of the output layer as the fourth constraint; and determining the first, second, third, and fourth constraints as model constraints.

[0009] Optionally, the operating cost of the distribution network can be obtained by determining the electricity purchase price and total electricity purchase volume of the distribution network, calculating the product of the electricity purchase price and total electricity purchase volume, and obtaining the operating cost of the distribution network.

[0010] Optionally, generating multiple candidate distribution network operation schemes through distribution network operation schemes includes: randomly generating multiple candidate distribution network operation schemes to obtain a set of candidate distribution network operation schemes; for each candidate distribution network operation scheme, calculating the sum of the energy storage power, generation power, and power purchase power of the distribution network of the candidate distribution network operation scheme to obtain the first load power of the candidate distribution network operation scheme; determining the second load power of the distribution network operation scheme, removing candidate distribution network operation schemes whose first load power differs from the second load power from the set of candidate distribution network operation schemes, and removing candidate distribution network operation schemes that do not meet the power constraint conditions from the set of candidate distribution network operation schemes to obtain an updated set of candidate distribution network operation schemes; for each candidate distribution network operation scheme in the updated set of candidate distribution network operation schemes, calculating the power difference between the candidate distribution network operation scheme and the distribution network operation scheme, and determining the candidate distribution network operation scheme with a power difference less than a power difference threshold as a candidate distribution network operation scheme, wherein the power difference includes at least one of the following: the energy storage power difference, the generation power difference, and the power purchase power difference of the distribution network.

[0011] Optionally, determining the constraint satisfaction probability of a distribution network operation scheme based on multiple undetermined distribution network operation schemes includes: calculating a first per-unit value and a second per-unit value for each undetermined distribution network operation scheme based on the target model; identifying undetermined distribution network operation schemes whose first per-unit value or second per-unit value is less than a preset value as constraint violation schemes, determining the number of constraint violation schemes to obtain a first number; determining the number of undetermined distribution network operation schemes to obtain a second number, and calculating the ratio of the first number to the second number to obtain the constraint satisfaction probability of the distribution network operation scheme.

[0012] Optionally, determining the target distribution network operation scheme from multiple undetermined distribution network operation schemes and distribution network operation schemes based on constraint satisfaction probability includes: if the constraint satisfaction probability of a distribution network operation scheme is less than a constraint satisfaction probability threshold, determining the distribution network operation scheme as the target distribution network operation scheme; if the constraint satisfaction probability of a distribution network operation scheme is greater than or equal to the constraint satisfaction probability threshold, using each undetermined distribution network operation scheme as an updated distribution network operation scheme, generating multiple sub-undetermined distribution network operation schemes through the updated distribution network operation schemes, and calculating the constraint satisfaction probability of the undetermined distribution network operation scheme based on the multiple sub-undetermined distribution network operation schemes and the undetermined distribution network operation schemes to obtain a set of constraint satisfaction probabilities; determining at least one undetermined distribution network operation scheme with a constraint satisfaction probability less than the constraint satisfaction probability threshold from multiple undetermined distribution network operation schemes, and determining the scheme with the minimum operating cost among the at least one undetermined distribution network operation schemes as the target distribution network operation scheme.

[0013] To achieve the above objectives, according to another aspect of this application, a device for acquiring a distribution network operation scheme is provided. The device includes: an acquisition unit, configured to acquire multiple sets of historical operation data within a preset period of the distribution network, and calculate a first per-unit value and a second per-unit value based on each set of historical operation data, wherein the first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network; and a conversion unit, configured to convert a target model into model constraint conditions, acquire the operation constraint conditions of the distribution network, and combine the model constraint conditions and the operation constraint conditions into a constraint condition set, wherein the target model is trained from multiple sets of training samples, each... The training samples include a set of historical operating data and a set of first and second per-unit values ​​of the historical operating data; the solution unit is used to solve the set of constraints based on the optimization objective of minimizing the operating cost of the distribution network to obtain the distribution network operation scheme, wherein the distribution network operation scheme includes at least the energy storage power, generation power and power purchase power of the distribution network; the generation unit is used to generate multiple undetermined distribution network operation schemes from the distribution network operation schemes, determine the constraint satisfaction probability of the distribution network operation scheme based on the multiple undetermined distribution network operation schemes, and determine the target distribution network operation scheme from the multiple undetermined distribution network operation schemes and the distribution network operation scheme based on the constraint satisfaction probability.

[0014] To achieve the above objectives, according to another aspect of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the method for obtaining the power distribution network operation scheme described in various embodiments of this application.

[0015] This application employs the following steps: acquiring multiple sets of historical operating data within a preset period of the distribution network; calculating a first per-unit value and a second per-unit value for each set of historical operating data, wherein the first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network; converting the target model into model constraints and acquiring the operating constraints of the distribution network; combining the model constraints and operating constraints into a constraint set, wherein the target model is trained from multiple sets of training samples, each set of training samples including a set of historical data. The system uses operational data and a set of historical operational data, including first and second per-unit values. With minimizing the operating cost of the distribution network as the optimization objective, it solves the constraint set based on the optimization objective to obtain a distribution network operation scheme. This scheme includes at least the distribution network's energy storage capacity, generation capacity, and power purchase capacity. Multiple undetermined distribution network operation schemes are generated from these schemes. The constraint satisfaction probability of each undetermined scheme is determined based on these schemes. Based on this probability, a target distribution network operation scheme is determined from the multiple undetermined schemes and the target scheme, thus solving the problem of low efficiency in obtaining real-time distribution network operation schemes in related technologies. A machine learning-based target model learns the distribution of historical operational data of the distribution network. This learned data distribution is then embedded into the constraint set for real-time operation optimization, replacing the power flow calculation part. Furthermore, considering the error in the neural network fitting in the target model, to improve economic efficiency while ensuring safety constraints are met, a Monte Carlo method is used after solving the optimization model to evaluate the constraint satisfaction probability and correct the distribution network operation scheme, thereby improving the efficiency of obtaining real-time distribution network operation schemes. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0017] Figure 1 This is a flowchart of a method for obtaining a power distribution network operation plan according to an embodiment of this application;

[0018] Figure 2 This is a schematic diagram of an optional power distribution network operation scheme acquisition method provided according to an embodiment of this application;

[0019] Figure 3 This is a schematic diagram of a device for obtaining a power distribution network operation plan according to an embodiment of this application;

[0020] Figure 4This is a schematic diagram of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0025] It should be noted that the information collected is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data all comply with the relevant laws, regulations and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation portals are provided for users to choose to authorize or refuse.

[0026] The present invention will now be described in conjunction with preferred implementation steps. Figure 1 This is a flowchart of a method for obtaining a power distribution network operation plan according to an embodiment of this application, such as... Figure 1 As shown, the method includes the following steps:

[0027] Step S101: Obtain multiple sets of historical operating data within a preset period of the distribution network, and calculate the first per-unit value and the second per-unit value of the historical operating data based on each set of historical operating data. The first per-unit value is the minimum per-unit value between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value between the line current and current boundary constraints of each line in the distribution network.

[0028] Specifically, historical operating data can include node voltages, line currents, and network losses (energy losses during distribution network operation) of each node in the distribution network. Historical operating data also includes parameters such as the safe operating range of node voltages (i.e., voltage boundary constraints) and the safe operating range of line currents (i.e., current boundary constraints). The preset period can be one year or one month. By acquiring multiple sets of historical operating data from the distribution network over the past year, and calculating the first per-unit value of the interval between the node voltage of each node and its corresponding voltage boundary constraint, and the second per-unit value of the interval between the line current of each line and its corresponding current boundary constraint, the operating data of each node in the distribution network is characterized by the first and second per-unit values. The distribution of the distribution network operating data is characterized by the first and second per-unit values ​​of all nodes.

[0029] Step S102: Convert the target model into model constraints and obtain the operation constraints of the distribution network. Combine the model constraints and operation constraints into a set of constraints. The target model is trained by multiple sets of training samples. Each set of training samples includes a set of historical operation data and a set of first and second per-unit values ​​of the historical operation data.

[0030] Specifically, as the topological complexity of distribution networks gradually increases, it becomes increasingly difficult to obtain detailed distribution network topology and impedance data. Therefore, a target model is trained using deep learning, and deep learning technology is used to learn the distribution of historical operating data of the distribution network. Furthermore, the learned data distribution is embedded into the real-time operation optimization of the distribution network, that is, the target model is converted into model constraints, and the model constraints replace the power flow calculation part of the distribution network operating data in related technologies.

[0031] For example, the target model can be three deep neural networks composed of multilayer sensing mechanisms, which learn the following in the distribution network under different load distributions (distribution of node voltage and line current) and different historical distribution network operation schemes: a. the minimum per-unit value of the interval between node voltage and its corresponding voltage boundary constraint among all nodes; b. the minimum per-unit value of the interval between line current and its corresponding current boundary constraint among all lines; c. the overall network loss rate of the distribution network.

[0032] After determining the target model, the neural network is decomposed to convert the target model into model constraints. Based on the obtained combination of operating constraints of the distribution network, the set of constraints that need to be satisfied in the real-time optimized operation scheme of the distribution network is obtained.

[0033] Step S103: Taking the minimization of the operating cost of the distribution network as the optimization objective, the set of constraints is solved based on the optimization objective to obtain the distribution network operation scheme. The distribution network operation scheme includes at least the energy storage capacity, power generation capacity, and power purchase capacity of the distribution network.

[0034] Specifically, after determining the set of constraints, the optimal distribution network operation scheme can be obtained by solving the set of constraints. The distribution network operation scheme can include the expected total energy storage capacity of all energy storage system nodes in the distribution network, the total power purchase capacity of all user nodes in the distribution network, and the total power generation capacity of all generation nodes in the distribution network. Solving the set of constraints yields multiple distribution network operation schemes. Minimizing the operating cost of the distribution network is selected as the optimization objective of the distribution network operation scheme, thereby solving for the distribution network operation scheme that minimizes the operating cost and satisfies the set of constraints.

[0035] Step S104: Generate multiple undetermined distribution network operation schemes through the distribution network operation schemes, determine the constraint satisfaction probability of the distribution network operation schemes based on the multiple undetermined distribution network operation schemes, and determine the target distribution network operation scheme from the multiple undetermined distribution network operation schemes and the distribution network operation schemes based on the constraint satisfaction probability.

[0036] Specifically, considering the errors in the distribution network operation data fitted by the neural network, in order to ensure that the distribution network operation scheme meets safety constraints (i.e., meets the set of constraints) while improving economic efficiency, the Monte Carlo method is used to generate multiple undetermined distribution network operation schemes after solving the optimization model, and the constraint satisfaction probability of each distribution network operation scheme is evaluated. Based on the constraint satisfaction probability, the distribution network operation scheme is further optimized, thereby obtaining the final optimized target distribution network operation scheme.

[0037] For example, after each distribution network operation scheme is solved, the Monte Carlo method is used to generate a set of suboptimal scheduling schemes (i.e., undetermined distribution network operation schemes) near the optimal scheduling scheme (i.e., the distribution network operation scheme itself). Then, a fully trained neural network is used to calculate the set of unsatisfied constraints for these undetermined distribution network operation schemes, and the constraint satisfaction probability of the distribution network operation scheme is calculated based on the undetermined distribution network operation schemes. If the constraint satisfaction probability exceeds a specified threshold, undetermined distribution network operation schemes that satisfy the constraint set are selected. The Monte Carlo method is repeated for these undetermined distribution network operation schemes to generate sub-undetermined distribution network operation schemes for each undetermined distribution network operation scheme, and the constraint satisfaction probability of each undetermined distribution network operation scheme that satisfies the constraint set is calculated based on the sub-undetermined distribution network operation schemes. Undetermined distribution network operation schemes whose constraint satisfaction probability does not exceed the threshold are identified, and the scheme that minimizes the operating cost is selected as the target distribution network operation scheme.

[0038] The method for obtaining a distribution network operation scheme provided in this application embodiment acquires multiple sets of historical operation data within a preset period of the distribution network. Based on each set of historical operation data, it calculates a first per-unit value and a second per-unit value of the historical operation data. The first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network. The method converts the target model into model constraints and acquires the operation constraints of the distribution network. The model constraints and operation constraints are combined into a constraint set. The target model is trained from multiple sets of training samples. This method includes a set of historical operating data, and first and second per-unit values ​​of that historical operating data. With minimizing the operating cost of the distribution network as the optimization objective, it solves the constraint set based on the optimization objective to obtain a distribution network operation scheme. This scheme includes at least the distribution network's energy storage capacity, generation capacity, and power purchase capacity. Multiple undetermined distribution network operation schemes are generated from these schemes. The constraint satisfaction probability of each undetermined scheme is determined based on these schemes. Based on this probability, a target distribution network operation scheme is determined from the multiple undetermined schemes and the current distribution network operation scheme. This solves the problem of low efficiency in obtaining real-time distribution network operation schemes in related technologies. Furthermore, by learning the distribution of historical operating data of the distribution network through a machine learning-based target model, the learned data distribution is further embedded into the constraint set for real-time operation optimization of the distribution network, replacing the distribution network power flow calculation part. Furthermore, considering the errors in the neural network fitting in the target model, in order to improve economic efficiency while ensuring that safety constraints are met as much as possible, the Monte Carlo method is used to evaluate the probability of constraint satisfaction after the model is optimized, and the distribution network operation scheme is corrected, thereby improving the efficiency of obtaining the distribution network operation scheme.

[0039] By calculating per-unit values ​​to characterize the historical operating data of the distribution network, optionally, in the method for obtaining the distribution network operation scheme provided in this application embodiment, calculating the first per-unit value and the second per-unit value of the historical operating data based on each set of historical operating data includes: determining the node voltage of each node and the line current of each line in the distribution network; determining the maximum node voltage and the minimum node voltage in the voltage boundary constraints of the distribution network; and determining the maximum line current in the current boundary constraints of the distribution network; for each node, calculating the first difference between the maximum node voltage and the node voltage of the node, and calculating the ratio of the first difference to the rated voltage of the distribution network, to obtain... The first voltage per-unit value of the node; the second difference between the node voltage and the minimum node voltage, and the ratio of the second difference to the rated voltage of the distribution network, to obtain the second voltage per-unit value of the node; the minimum voltage per-unit value is determined from the first voltage per-unit value and the second voltage per-unit value of all nodes, and the minimum voltage per-unit value is determined as the first voltage per-unit value; for each line, the third difference between the maximum line current and the line current, and the ratio of the third difference to the maximum line current, to obtain the current per-unit value of the line; the minimum current per-unit value is determined from the current per-unit values ​​of all lines, and the minimum current per-unit value is determined as the second voltage per-unit value.

[0040] Specifically, firstly, operational data such as node voltage and line current of each node in the distribution network are obtained, along with parameters such as the safe operating range of node voltage and line current, i.e., the maximum node voltage, minimum node voltage, and maximum line current. The formula for calculating the first per-unit value is as follows:

[0041]

[0042] Where, ΔV t It is the minimum per-unit value of the node voltage of all nodes in the distribution network at time t and its voltage boundary constraint interval, that is, the first per-unit value, V. i,t It is the node voltage of node i in the distribution network at time t. This is the rated voltage of node i (e.g., 220V). It is the maximum node voltage of the voltage boundary constraints for the safe operation of the distribution network. It is the minimum node voltage under voltage boundary constraints. The formula for calculating the second per-unit value is as follows:

[0043]

[0044] Where, ΔI t It is the minimum per-unit value of the line current and its current boundary constraint interval for all lines in the distribution network at time t, also known as the second per-unit value, I. j,tIt is the line current of line j in the distribution network at time t. It is the rated line current of line j, that is, the maximum line current in the current boundary constraints.

[0045] This embodiment characterizes the operation of the distribution network by calculating the first per-unit value and the second per-unit value. It does not rely on the detailed network structure of the distribution network and only requires historical operation data of the distribution network to predict the safe operation of the distribution network, thereby improving the efficiency of obtaining distribution network operation plans.

[0046] To find a safe operating scheme for the distribution network, the target model learned from historical operating data needs to be converted into model constraints. Optionally, in the method for obtaining the distribution network operating scheme provided in this application embodiment, converting the target model into model constraints includes: determining the input layer, hidden layer, and output layer of the target model, and determining the model parameters of the target model; determining the expression of the input layer as the first constraint, the linear mapping expression of the hidden layer as the second constraint, the output expression of the activation function of the hidden layer as the third constraint, and the expression of the output layer as the fourth constraint; and determining the first, second, third, and fourth constraints as model constraints.

[0047] In some examples, the target model learned from historical operating data in this embodiment may include three: a neural network model describing voltage constraint mapping, current constraint mapping, and distribution network loss mapping. The expressions for the above three neural network models are as follows:

[0048]

[0049] Where, π V Θ V D represents the neural network describing the voltage constraint mapping relationship and its network parameters. t P t These represent the load distribution and adjustable resource scheduling plan of the distribution network at time t. The load distribution includes, for example, the node voltage of each node and the line current of each line. The adjustable resource scheduling plan is the distribution network operation scheme corresponding to the historical operating data of the distribution network, such as the energy storage capacity, generation capacity, and power purchase capacity of the distribution network. π I Θ I These represent the neural network describing the current constraint mapping relationship and its network parameters, π. L Θ L These represent the neural network describing the mapping relationship of distribution network losses and its network parameters, respectively.

[0050] Furthermore, the multilayer perceptron of the target model adopts a standard structure, and the specific model constraints for transforming the target model are as follows:

[0051]

[0052] Where, x t The input data is D in formulas (3), (4), and (5). t P t , It is the input layer. It is the output of the linear mapping of hidden layer l. The output of the activation function of hidden layer l. The output of the activation function of the l-1 hidden layer; L is the set of hidden layers, and the parameters (W) l b l ) and (w |L|+1 b |L|+1 ) represents the weights and biases of the objective function, h t It is the output of the output layer of the multilayer perceptron neural network. Formula (6) is also the first constraint, formula (7) is also the second constraint, and formula (9) is also the fourth constraint.

[0053] Since equation (8) is a nonlinear constraint, while the others are linear constraints, the constraint decomposition method of operations research is used to decompose the nonlinear term in equation (8) into auxiliary standard linear constraints containing 0-1 variables:

[0054]

[0055] in, As an auxiliary variable, M is a huge constant. Let it be a 0-1 variable. In equation (10), when When it is a positive number, when When it is negative, Formula (10) is also the third constraint condition. By using the operation research constraint decomposition method, the nonlinear neurons in the above neural network are decomposed into auxiliary standard linear constraints with 0-1 variables, and added to the distribution network optimization operation constraints without power flow calculation, forming a complete distribution network optimization problem.

[0056] This embodiment uses a neural network to record historical operating data of the distribution network, avoiding detailed distribution network modeling. This avoids the problem of being unable to model due to opaque topology data, reduces the complexity of optimization modeling, and improves the efficiency of optimizing and solving distribution network operation schemes.

[0057] Furthermore, after determining the model constraints, it is also necessary to obtain the operational constraints of the distribution network. The operational constraints may include power balance constraints, energy storage output constraints, energy storage capacity constraints, gas turbine output constraints, voltage safety constraints, current safety constraints, etc. The operational constraints, the neural network calculation equation constraints of voltage / current / network loss represented by formulas (3)-(5), and the model constraints of formulas (6), (7), (9), and (10) are combined to form the constraint set. The expression of the constraint set is as follows:

[0058]

[0059] Where Di,t represents the electrical load of node i at time t, and Pi,t represents the power generation of node i at time t. It represents the photovoltaic power generation of node i at time t. It is the energy storage charging and discharging power of node i at time t. It represents the power generation of the gas turbine unit at node i at time t. It represents the amount of electricity purchased by node i from the upper-level power grid at time t. It refers to the state of charge of energy storage devices. It is the upper limit of energy storage capacity. This is the maximum charging power of the energy storage. It is the maximum discharge power of the energy storage. It is the maximum discharge power of the gas turbine unit.

[0060] After determining the set of constraints, the set of constraints is solved by minimizing the operating cost to obtain the distribution network operation scheme. Optionally, in the method for obtaining the distribution network operation scheme provided in the embodiments of this application, the operating cost of the distribution network is obtained by: determining the electricity purchase price and total electricity purchase volume of the distribution network, calculating the product of the electricity purchase price and the total electricity purchase volume, and obtaining the operating cost of the distribution network.

[0061] Specifically, the formula for calculating the operating cost of a power distribution network is as follows:

[0062]

[0063] Among them, c t It is the electricity purchase price of the distribution network at each moment. It represents the total electricity purchased at any given moment.

[0064] This embodiment obtains the optimal operating scheme of the distribution network at this time by solving the optimization problem consisting of a set of constraints with the goal of minimizing operating costs.

[0065] To improve the robustness of distribution network operation schemes, multiple undetermined distribution network operation schemes are generated to optimize the distribution network operation scheme. Optionally, in the method for obtaining distribution network operation schemes provided in this application embodiment, generating multiple undetermined distribution network operation schemes through distribution network operation schemes includes: randomly generating multiple candidate distribution network operation schemes to obtain a set of candidate distribution network operation schemes; for each candidate distribution network operation scheme, calculating the sum of the energy storage power, generation power, and power purchase power of the distribution network of the candidate distribution network operation scheme to obtain the first load power of the candidate distribution network operation scheme; determining the second load power of the distribution network operation scheme, and combining the first load power with the second load power. Candidate distribution network operation schemes with different power are removed from the candidate distribution network operation scheme set, and candidate distribution network operation schemes that do not meet the power constraints are also removed from the candidate distribution network operation scheme set, resulting in an updated candidate distribution network operation scheme set. For each candidate distribution network operation scheme in the updated candidate distribution network operation scheme set, the power difference between the candidate distribution network operation scheme and the distribution network operation scheme is calculated. Candidate distribution network operation schemes with power difference less than the power difference threshold are determined as undetermined distribution network operation schemes. The power difference includes at least one of the following: the energy storage power difference of the distribution network, the power generation power difference, and the power purchase power difference.

[0066] Specifically, in solving the distribution network operation scheme Then, the expression for the distribution network operation scheme is as follows:

[0067]

[0068] in, For the energy storage capacity in the distribution network operation plan, The power generation capacity in the power distribution network operation plan, Let represent the power purchase capacity in the distribution network operation plan. Using the Monte Carlo method, n sets of candidate distribution network operation plans are generated near the current operation plan. The expressions for these candidate plans are as follows:

[0069]

[0070] Among them, P n,t For the nth candidate distribution network operation scheme, Let n be the energy storage capacity in the nth candidate distribution network operation scheme. Let n be the power generation capacity in the nth candidate distribution network operation scheme. Let P be the power purchase capacity in the nth candidate distribution network operation scheme. The undetermined distribution network operation schemes selected from the candidate schemes need to meet certain constraints. n,t The load balance constraint of the distribution network should be satisfied. The expression for the load balance constraint is as follows:

[0071]

[0072] In formula (15), the left side represents the second load power, and the right side represents the first load power. The generated undetermined distribution network operation scheme P n,t Basic unit output constraints must also be met:

[0073]

[0074] Among them, P ES,c It is the maximum charging power of energy storage, P ES,d It is the maximum discharge power of the energy storage, P GT,max The maximum discharge power of the gas turbine unit generates the undetermined power distribution network operation scheme P. n,t It should also meet the requirements of unit output changes and distribution network operation plans. The maximum difference between them is the power difference threshold λ, and the power difference constraint expression is as follows:

[0075]

[0076] Since the undetermined distribution network operation plan changes very little compared to the original distribution network operation plan, network losses are considered... The constraints remain unchanged. Therefore, the constraints of the distribution network are not considered.

[0077] This embodiment generates multiple undetermined distribution network operation schemes to optimize the distribution network operation scheme, thereby improving the robustness of the distribution network operation scheme.

[0078] To select a more robust target distribution network operation scheme from the distribution network operation schemes and pending distribution network operation schemes, it is necessary to calculate the constraint satisfaction probability of each scheme. Optionally, in the distribution network operation scheme acquisition method provided in this application embodiment, determining the constraint satisfaction probability of the distribution network operation scheme based on multiple pending distribution network operation schemes includes: calculating a first per-unit value and a second per-unit value for each pending distribution network operation scheme based on the target model; identifying pending distribution network operation schemes whose first per-unit value or second per-unit value is less than a preset value as constraint violation schemes, determining the number of constraint violation schemes to obtain a first number; determining the number of pending distribution network operation schemes to obtain a second number, and calculating the ratio of the first number to the second number to obtain the constraint satisfaction probability of the distribution network operation scheme.

[0079] Specifically, the first and second per-unit values ​​for each pending distribution network operation scheme are calculated using the following formulas:

[0080]

[0081] By inputting the proposed distribution network operation schemes into formulas (18) and (19), the first and second per-unit values ​​of the proposed distribution network operation schemes are obtained. After obtaining the first and second per-unit values ​​of all proposed distribution network operation schemes, the voltage / current constraint satisfaction of each proposed distribution network operation scheme is calculated.

[0082]

[0083] in, A value of 1 indicates a violation of the constraint scheme, that is, a scheme with potential safety hazards. Then, based on the violation status of all pending distribution network operation schemes, the distribution network operation scheme is calculated. The probability η of satisfying the neighborhood constraint t The formula for calculating the probability of constraint satisfaction is as follows:

[0084]

[0085] in, Let N represent the first quantity and N represent the second quantity.

[0086] This embodiment selects a more robust target distribution network operation scheme from the distribution network operation schemes and undetermined distribution network operation schemes by calculating the constraint satisfaction probability of each scheme.

[0087] After calculating the constraint satisfaction probability of the distribution network operation scheme, a target distribution network operation scheme is selected from the distribution network operation schemes and the undetermined distribution network operation schemes. Optionally, in the method for obtaining the distribution network operation scheme provided in this application embodiment, determining the target distribution network operation scheme from multiple undetermined distribution network operation schemes and the distribution network operation scheme based on the constraint satisfaction probability includes: if the constraint satisfaction probability of the distribution network operation scheme is less than the constraint satisfaction probability threshold, the distribution network operation scheme is determined as the target distribution network operation scheme; if the constraint satisfaction probability of the distribution network operation scheme is greater than or equal to the constraint satisfaction probability threshold, each undetermined distribution network operation scheme is used as the updated distribution network operation scheme, multiple sub-undetermined distribution network operation schemes are generated through the updated distribution network operation schemes, and the constraint satisfaction probability of the undetermined distribution network operation scheme is calculated based on the multiple sub-undetermined distribution network operation schemes and the undetermined distribution network operation schemes to obtain a set of constraint satisfaction probabilities; at least one undetermined distribution network operation scheme with a constraint satisfaction probability less than the constraint satisfaction probability threshold is determined from the multiple undetermined distribution network operation schemes, and the scheme with the minimum operating cost among the at least one undetermined distribution network operation scheme is determined as the target distribution network operation scheme.

[0088] Specifically, when the constraint is satisfied with probability η t The probability threshold for satisfying the less than constraint is met. If the proposed distribution network operation plan meets the operational safety requirements, then that plan becomes the final target distribution network operation plan. Otherwise, if the proposed plan does not meet the operational safety requirements, a target distribution network operation plan needs to be selected from the pool of pending distribution network operation plans.

[0089] For example, among multiple undetermined distribution network operation schemes, the selection... The constraints do not violate the rules, that is P, the pending distribution network operation plan m,t Where the subscript m represents the number of the undetermined distribution network operation scheme that satisfies the constraints. For all undetermined distribution network operation schemes P... m,t Multiple undetermined sub-distribution network operation schemes are generated, and the final calculation of these undetermined distribution network operation schemes P is performed. m,t The probability η of satisfying the constraint m,t Find it. P, the pending distribution network operation plan m,t And select the undetermined distribution network operation scheme P that minimizes operating costs. m,t This scheme is the final target distribution network operation scheme.

[0090] If no undetermined distribution network operation scheme satisfies the constraints, multiple sub-undetermined distribution network operation schemes are repeatedly generated for these undetermined distribution network operation schemes. Through iterative tree retrieval, undetermined distribution network operation schemes with a probability of satisfying constraints less than the constraint satisfaction threshold are searched until a scheme with a constraint satisfaction probability less than the constraint satisfaction probability threshold is found, which is then the final target distribution network operation scheme.

[0091] This embodiment overcomes the constraint violations caused by the uncertainty of neural network fitting to a certain extent by using the Monte Carlo-based neighborhood constraint violation posterior, and ensures the minimization of the operating cost of the target distribution network operation scheme.

[0092] According to another embodiment of this application, a method for obtaining an optional power distribution network operation scheme is also provided. Figure 2 This is a schematic diagram illustrating a method for obtaining optional power distribution network operation schemes according to embodiments of this application. For example... Figure 2 As shown, the scheme includes:

[0093] Step 01: Obtain the distribution network operation data and design operating range, and calculate the distance between the distribution network operation data and the safety boundary and the distribution network loss at different time points.

[0094] Specifically, the distribution network operation data can include the node voltage, line current, network loss and other operation data of each node in the distribution network. The safety boundary can be voltage boundary constraint and current boundary constraint. The distance between the distribution network operation data and the safety boundary is also known as the first per-unit value and the second per-unit value.

[0095] Step 02: Construct three neural networks to learn the distribution network voltage safety margin, current safety margin, and distribution network loss under different conditions.

[0096] Specifically, the three neural networks are the networks corresponding to formulas (3), (4) and (5). Different situations refer to the situations corresponding to each group of historical running data of different training samples. The distribution network voltage safety constraint margin is also the first per-unit value, and the current safety constraint margin is also the second per-unit value.

[0097] Step 03: Decompose the nonlinear neurons in the neural network into auxiliary standard linear constraints and add them to the running optimization constraints.

[0098] Specifically, the nonlinear neuron is also known as formula (8), the auxiliary standard linear constraint is also known as formula (10), and the set of constraint conditions obtained after adding it to the running optimization constraint is also known as formula (12).

[0099] Step 04: Monte Carlo generates suboptimal scheduling schemes near the optimal scheduling scheme and calculates the probability of neighborhood constraint violation in all cases.

[0100] Specifically, the optimal scheduling scheme is also known as the distribution network operation scheme, and the suboptimal scheduling scheme is also known as the undetermined distribution network operation scheme. All cases refer to the cases corresponding to different suboptimal scheduling schemes, and the neighborhood constraint default probability is also known as the constraint satisfaction probability.

[0101] Step 05: If the probability of violating the neighborhood constraint exceeds the specified value, repeat step 04 for the suboptimal scheduling plan that does not violate the constraint. Iterate continuously through tree search until a scheduling plan that satisfies the constraint is found.

[0102] Specifically, the specified value, also known as the constraint satisfaction probability threshold, refers to a suboptimal dispatch plan that does not violate the constraint and whose first and second per-unit values ​​are both greater than or equal to 0. A dispatch plan that satisfies the constraints refers to a target distribution network operation scheme where the constraint satisfaction probability is less than the constraint satisfaction probability threshold.

[0103] This embodiment utilizes a method for obtaining optional distribution network operation schemes, employing neural networks to record historical data. This avoids detailed modeling of the distribution network, thus preventing the modeling problem caused by opaque topology data. Furthermore, by omitting distribution network power flow calculations, the complexity of the optimization model is reduced, improving the efficiency of the optimization solution. The use of a Monte Carlo-based posterior regression for neighborhood constraint violations overcomes, to some extent, the constraint violations caused by the uncertainty of neural network fitting, ensuring the economic efficiency of the optimized scheduling scheme. Although the iterative process is complex, the extremely fast computer solution speed, due to the simple random data generation / neural network data generation process, does not affect the algorithm's overall efficiency.

[0104] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0105] This application also provides a device for obtaining a distribution network operation plan. It should be noted that this device can be used to execute the method for obtaining a distribution network operation plan provided in this application. The following describes the device for obtaining a distribution network operation plan provided in this application.

[0106] Figure 3 This is a schematic diagram of a device for obtaining a power distribution network operation plan according to an embodiment of this application. Figure 3 As shown, the device includes:

[0107] The acquisition unit 301 is used to acquire multiple sets of historical operating data within a preset period of the distribution network, and calculate the first per-unit value and the second per-unit value of the historical operating data based on each set of historical operating data. The first per-unit value is the minimum per-unit value between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value between the line current and current boundary constraints of each line in the distribution network.

[0108] The conversion unit 302 is used to convert the target model into model constraints and obtain the operation constraints of the distribution network. The model constraints and operation constraints are combined into a set of constraints. The target model is trained by multiple sets of training samples. Each set of training samples includes a set of historical operation data and a first per-unit value and a second per-unit value of the historical operation data.

[0109] The solver unit 303 is used to solve the set of constraints based on the optimization objective of minimizing the operating cost of the distribution network, and obtain the distribution network operation scheme. The distribution network operation scheme includes at least the energy storage capacity, power generation capacity and power purchase capacity of the distribution network.

[0110] The generation unit 304 is used to generate multiple pending distribution network operation schemes through the distribution network operation scheme, determine the constraint satisfaction probability of the distribution network operation scheme based on the multiple pending distribution network operation schemes, and determine the target distribution network operation scheme from the multiple pending distribution network operation schemes and the distribution network operation scheme based on the constraint satisfaction probability.

[0111] The distribution network operation scheme acquisition device provided in this application embodiment acquires multiple sets of historical operation data within a preset period of the distribution network through acquisition unit 301, and calculates a first per-unit value and a second per-unit value of the historical operation data based on each set of historical operation data. The first per-unit value is the minimum per-unit value between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value between the line current and current boundary constraints of each line in the distribution network. Conversion unit 302 converts the target model into model constraint conditions and acquires the operation constraint conditions of the distribution network. It combines the model constraint conditions and the operation constraint conditions into a constraint condition set. The target model is trained from multiple sets of training samples, each set of training samples including a set of historical operation data and a set of first and second per-unit values ​​of the historical operation data. Solving unit 303... Minimizing the operating cost of the distribution network is the optimization objective. Based on this objective, the constraint set is solved to obtain a distribution network operation plan. This plan includes at least the distribution network's energy storage capacity, generation capacity, and power purchase capacity. Generation unit 304 generates multiple undetermined distribution network operation plans from these plans. Based on these plans, the constraint satisfaction probability is determined. Then, based on this probability, a target distribution network operation plan is selected from the undetermined and target plans. This addresses the low efficiency of obtaining real-time distribution network operation plans in related technologies. A machine learning-based target model learns the distribution of historical operation data of the distribution network and embeds this data distribution into the constraint set for real-time operation optimization, replacing the power flow calculation part. Furthermore, considering the fitting error in the neural network of the target model, to improve economic efficiency while ensuring safety constraints are met, a Monte Carlo method is used after solving the optimization model to evaluate the constraint satisfaction probability and correct the distribution network operation plan, thereby improving the efficiency of obtaining the distribution network operation plan.

[0112] Optionally, in the distribution network operation scheme acquisition device provided in this application embodiment, the acquisition unit 301 includes: a first determining module, used to determine the node voltage of each node and the line current of each line in the distribution network, determine the maximum node voltage and the minimum node voltage in the voltage boundary constraints of the distribution network, and determine the maximum line current in the current boundary constraints of the distribution network; a first calculating module, used to calculate, for each node, a first difference between the maximum node voltage and the node voltage of the node, and calculate the ratio of the first difference to the rated voltage of the distribution network to obtain the first voltage per unit value of the node; a second calculating module, used to calculate the ratio of the node voltage of the node to the minimum node voltage of the node. The second voltage per-unit value of the node is obtained by calculating the second voltage difference and the ratio of the second voltage difference to the rated voltage of the distribution network; the second determination module is used to determine the minimum voltage per-unit value from the first voltage per-unit value and the second voltage per-unit value of all nodes, and to determine the minimum voltage per-unit value as the first per-unit value; the third calculation module is used to calculate the third difference between the maximum line current and the line current for each line, and to calculate the ratio of the third difference to the maximum line current, to obtain the current per-unit value of the line; the third determination module is used to determine the minimum current per-unit value from the current per-unit values ​​of all lines, and to determine the minimum current per-unit value as the second per-unit value.

[0113] Optionally, in the distribution network operation scheme acquisition device provided in the embodiments of this application, the conversion unit 302 includes: a fourth determining module, used to determine the input layer, hidden layer and output layer of the target model, and to determine the model parameters of the target model; a fifth determining module, used to determine the expression of the input layer as the first constraint condition, the linear mapping expression of the hidden layer as the second constraint condition, the output expression of the activation function of the hidden layer as the third constraint condition, and the expression of the output layer as the fourth constraint condition; and a sixth determining module, used to determine the first constraint condition, the second constraint condition, the third constraint condition and the fourth constraint condition as model constraints.

[0114] Optionally, in the distribution network operation scheme acquisition device provided in the embodiments of this application, the operating cost of the distribution network is obtained by: determining the electricity purchase price and total electricity purchase volume of the distribution network, calculating the product of the electricity purchase price and the total electricity purchase volume, and obtaining the operating cost of the distribution network.

[0115] Optionally, in the distribution network operation scheme acquisition device provided in this application embodiment, the generation unit 304 includes: a first generation module, used to randomly generate multiple candidate distribution network operation schemes to obtain a set of candidate distribution network operation schemes; a fourth calculation module, used to calculate the sum of the energy storage power, power generation power, and power purchase power of the distribution network for each candidate distribution network operation scheme to obtain the first load power of the candidate distribution network operation scheme; and a seventh determination module, used to determine the second load power of the distribution network operation scheme and remove candidate distribution network operation schemes whose first load power and second load power are different from the candidate distribution network operation schemes. The system removes candidate distribution network operation schemes from the set of power grid operation schemes and removes candidate distribution network operation schemes that do not meet the power constraints from the set of candidate distribution network operation schemes, resulting in an updated set of candidate distribution network operation schemes. The fifth calculation module is used to calculate the power difference between the candidate distribution network operation scheme and the existing distribution network operation scheme for each candidate distribution network operation scheme in the updated set of candidate distribution network operation schemes. Candidate distribution network operation schemes with power differences less than the power difference threshold are identified as pending distribution network operation schemes. The power difference includes at least one of the following: the energy storage power difference of the distribution network, the power generation power difference, and the power purchase power difference.

[0116] Optionally, in the distribution network operation scheme acquisition device provided in this application embodiment, the generation unit 304 includes: a sixth calculation module, used to calculate a first per-unit value and a second per-unit value for each pending distribution network operation scheme based on the target model; an eighth determination module, used to determine the pending distribution network operation scheme with a first per-unit value or a second per-unit value less than a preset value as a constraint violation scheme, determine the number of constraint violation schemes, and obtain a first number; and a ninth determination module, used to determine the number of pending distribution network operation schemes, obtain a second number, calculate the ratio of the first number to the second number, and obtain the constraint satisfaction probability of the distribution network operation scheme.

[0117] Optionally, in the distribution network operation scheme acquisition device provided in this application embodiment, the generation unit 304 includes: a tenth determining module, used to determine the distribution network operation scheme as the target distribution network operation scheme when the constraint satisfaction probability of the distribution network operation scheme is less than the constraint satisfaction probability threshold; a second generating module, used to take each pending distribution network operation scheme as the updated distribution network operation scheme when the constraint satisfaction probability of the distribution network operation scheme is greater than or equal to the constraint satisfaction probability threshold, generate multiple sub-pending distribution network operation schemes through the updated distribution network operation schemes, and calculate the constraint satisfaction probability of the pending distribution network operation scheme based on the multiple sub-pending distribution network operation schemes and the pending distribution network operation schemes to obtain a set of constraint satisfaction probabilities; and an eleventh determining module, used to determine at least one pending distribution network operation scheme with a constraint satisfaction probability less than the constraint satisfaction probability threshold from the multiple pending distribution network operation schemes, and determine the scheme with the minimum operating cost among the at least one pending distribution network operation schemes as the target distribution network operation scheme.

[0118] The device for obtaining the power distribution network operation plan includes a processor and a memory. The aforementioned acquisition unit 301, conversion unit 302, solution unit 303, and generation unit 304 are all stored in the memory as program units. The processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0119] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters can improve the efficiency of obtaining power distribution network operation plans.

[0120] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0121] This invention provides a computer-readable storage medium storing a program that, when executed by a processor, implements a method for obtaining a power distribution network operation scheme.

[0122] This invention provides a processor for running a program, wherein the program executes a method for obtaining a power distribution network operation plan during runtime.

[0123] Figure 4 This is a schematic diagram of an electronic device provided according to an embodiment of this application. For example... Figure 4As shown, the electronic device 401 includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs the following steps: acquiring multiple sets of historical operating data within a preset period of the distribution network; calculating a first per-unit value and a second per-unit value of the historical operating data based on each set of historical operating data, wherein the first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network; converting the target model into model constraint conditions, acquiring the operating constraint conditions of the distribution network, and combining the model constraint conditions and the operating constraint conditions into constraint clauses. The system comprises a set of components, wherein the target model is trained from multiple sets of training samples. Each set of training samples includes a set of historical operating data and a set of first and second per-unit values ​​of the historical operating data. The optimization objective is to minimize the operating cost of the distribution network. Based on the optimization objective, the system solves for the set of constraints to obtain a distribution network operation scheme. Each distribution network operation scheme includes at least the energy storage capacity, generation capacity, and power purchase capacity of the distribution network. Multiple undetermined distribution network operation schemes are generated from these schemes. The constraint satisfaction probability of each undetermined scheme is determined based on these multiple undetermined schemes. Finally, the target distribution network operation scheme is determined from these undetermined schemes and the current distribution network operation scheme based on the constraint satisfaction probability. The devices used in this paper can be servers, PCs, tablets, mobile phones, etc.

[0124] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program with the following initialization steps: acquiring multiple sets of historical operating data within a preset period of the distribution network; calculating a first per-unit value and a second per-unit value of the historical operating data based on each set of historical operating data, wherein the first per-unit value is the minimum per-unit value of the interval between the node voltage and voltage boundary constraints of each node in the distribution network, and the second per-unit value is the minimum per-unit value of the interval between the line current and current boundary constraints of each line in the distribution network; converting the target model into model constraints, acquiring the operating constraints of the distribution network, and combining the model constraints and operating constraints into a constraint set. The target model is trained from multiple sets of training samples. Each set of training samples includes a set of historical operating data and a set of first and second per-unit values ​​of the historical operating data. The optimization objective is to minimize the operating cost of the distribution network. Based on the optimization objective, the set of constraints is solved to obtain the distribution network operation scheme. The distribution network operation scheme includes at least the energy storage capacity, power generation capacity, and power purchase capacity of the distribution network. Multiple undetermined distribution network operation schemes are generated from the distribution network operation schemes. The constraint satisfaction probability of the distribution network operation scheme is determined based on the multiple undetermined distribution network operation schemes. Based on the constraint satisfaction probability, the target distribution network operation scheme is determined from the multiple undetermined distribution network operation schemes and the distribution network operation scheme.

[0125] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0127] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0128] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0129] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0130] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0131] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0132] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0133] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0134] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for obtaining a power distribution network operation scheme, characterized in that, include: Multiple sets of historical operating data within a preset period of the distribution network are acquired, and a first per-unit value and a second per-unit value of the historical operating data are calculated based on each set of historical operating data. The first per-unit value is the minimum per-unit value between the node voltage and voltage boundary constraints of each node of the distribution network, and the second per-unit value is the minimum per-unit value between the line current and current boundary constraints of each line of the distribution network. The target model is converted into model constraints, and the operation constraints of the distribution network are obtained. The model constraints and the operation constraints are combined into a set of constraints. The target model is trained by multiple sets of training samples. Each set of training samples includes a set of historical operation data, a first per-unit value and a second per-unit value of the set of historical operation data. With the goal of minimizing the operating cost of the distribution network, the set of constraints is solved based on the goal to obtain a distribution network operation scheme, wherein the distribution network operation scheme includes at least the energy storage capacity, power generation capacity, and power purchase capacity of the distribution network. Multiple undetermined distribution network operation schemes are generated through the distribution network operation scheme. The constraint satisfaction probability of the distribution network operation scheme is determined based on the multiple undetermined distribution network operation schemes. The target distribution network operation scheme is determined from the multiple undetermined distribution network operation schemes and the distribution network operation scheme based on the constraint satisfaction probability. The calculation of the first and second per-unit values ​​of the historical operating data based on each set of historical operating data includes: determining the node voltage of each node and the line current of each line in the distribution network; determining the maximum and minimum node voltages in the voltage boundary constraints of the distribution network; and determining the maximum line current in the current boundary constraints of the distribution network; for each node, calculating the first difference between the maximum node voltage and the node voltage of the node, and calculating the ratio of the first difference to the rated voltage of the distribution network to obtain the first voltage per-unit value of the node; calculating the ratio of the node voltage of the node to the minimum node voltage... The second voltage difference is calculated, and the ratio of the second voltage difference to the rated voltage of the distribution network is calculated to obtain the second voltage per-unit value of the node; the minimum voltage per-unit value is determined from the first voltage per-unit value and the second voltage per-unit value of all nodes, and the minimum voltage per-unit value is determined as the first per-unit value; for each line, the third difference between the maximum line current and the line current of the line is calculated, and the ratio of the third difference to the maximum line current is calculated to obtain the current per-unit value of the line; the minimum current per-unit value is determined from the current per-unit values ​​of all lines, and the minimum current per-unit value is determined as the second per-unit value.

2. The method according to claim 1, characterized in that, Converting the target model into model constraints includes: The input layer, hidden layer, and output layer of the target model are determined, and the model parameters of the target model are also determined. The expression of the input layer is determined as the first constraint, the linear mapping expression of the hidden layer is determined as the second constraint, the output expression of the activation function of the hidden layer is determined as the third constraint, and the expression of the output layer is determined as the fourth constraint. The first constraint, the second constraint, the third constraint, and the fourth constraint are determined as the model constraints.

3. The method according to claim 1, characterized in that, The operating cost of the distribution network is obtained in the following way: Determine the electricity purchase price and total electricity purchase volume of the distribution network, calculate the product of the electricity purchase price and the total electricity purchase volume, and obtain the operating cost of the distribution network.

4. The method according to claim 1, characterized in that, The generation of multiple undetermined distribution network operation schemes through the aforementioned distribution network operation scheme includes: Multiple candidate distribution network operation schemes are randomly generated to obtain a set of candidate distribution network operation schemes; For each candidate distribution network operation scheme, the sum of the energy storage power, power generation power and power purchase power of the distribution network in the candidate distribution network operation scheme is calculated to obtain the first load power of the candidate distribution network operation scheme; Determine the second load power of the distribution network operation scheme, remove candidate distribution network operation schemes whose first load power is different from the second load power from the candidate distribution network operation scheme set, and remove candidate distribution network operation schemes that do not meet the power constraint conditions from the candidate distribution network operation scheme set to obtain an updated candidate distribution network operation scheme set; For each candidate distribution network operation scheme in the updated set of candidate distribution network operation schemes, the power difference between the candidate distribution network operation scheme and the distribution network operation scheme is calculated. The candidate distribution network operation scheme with the power difference less than the power difference threshold is determined as the undetermined distribution network operation scheme. The power difference includes at least one of the following: the energy storage power difference, the power generation power difference, and the power purchase power difference of the distribution network.

5. The method according to claim 1, characterized in that, The probability of constraint satisfaction of the distribution network operation scheme is determined based on the multiple undetermined distribution network operation schemes, including: Calculate the first and second per-unit values ​​for each undetermined distribution network operation scheme based on the target model; The undetermined distribution network operation scheme with the first per-unit value or the second per-unit value being less than a preset value is identified as a constraint violation scheme, and the number of the constraint violation schemes is determined to obtain a first number; The number of undetermined distribution network operation schemes is determined to obtain a second number. The ratio of the first number to the second number is calculated to obtain the constraint satisfaction probability of the distribution network operation scheme.

6. The method according to claim 1, characterized in that, Determining the target distribution network operation scheme from the plurality of undetermined distribution network operation schemes and the distribution network operation schemes based on the constraint satisfaction probability includes: If the probability of satisfying the constraints of the power distribution network operation scheme is less than the constraint satisfaction probability threshold, the power distribution network operation scheme shall be determined as the target power distribution network operation scheme. When the constraint satisfaction probability of the distribution network operation scheme is greater than or equal to the constraint satisfaction probability threshold, each undetermined distribution network operation scheme is used as the updated distribution network operation scheme. Multiple sub-undetermined distribution network operation schemes are generated through the updated distribution network operation schemes. The constraint satisfaction probability of the undetermined distribution network operation scheme is calculated based on the multiple sub-undetermined distribution network operation schemes and the undetermined distribution network operation scheme, and a set of constraint satisfaction probabilities is obtained. From the plurality of undetermined distribution network operation schemes, at least one undetermined distribution network operation scheme with a constraint satisfaction probability less than the constraint satisfaction probability threshold is determined, and the scheme with the minimum operating cost among the at least one undetermined distribution network operation schemes is determined as the target distribution network operation scheme.

7. A device for obtaining a power distribution network operation plan, characterized in that, include: The acquisition unit is used to acquire multiple sets of historical operating data within a preset period of the distribution network, and calculate a first per-unit value and a second per-unit value of the historical operating data based on each set of historical operating data, wherein the first per-unit value is the minimum per-unit value between the node voltage and voltage boundary constraints of each node of the distribution network, and the second per-unit value is the minimum per-unit value between the line current and current boundary constraints of each line of the distribution network. The conversion unit is used to convert the target model into model constraints and obtain the operation constraints of the distribution network. The model constraints and the operation constraints are combined into a set of constraints. The target model is trained by multiple sets of training samples. Each set of training samples includes a set of historical operation data, a first per-unit value and a second per-unit value of the set of historical operation data. The solution unit is used to solve the set of constraints based on the optimization objective of minimizing the operating cost of the distribution network to obtain the distribution network operation scheme, wherein the distribution network operation scheme includes at least the energy storage capacity, power generation capacity and power purchase capacity of the distribution network; The generation unit is configured to generate multiple undetermined distribution network operation schemes through the distribution network operation scheme, determine the constraint satisfaction probability of the distribution network operation scheme based on the multiple undetermined distribution network operation schemes, and determine the target distribution network operation scheme from the multiple undetermined distribution network operation schemes and the distribution network operation scheme based on the constraint satisfaction probability. The acquisition unit includes: a first determining module, used to determine the node voltage of each node and the line current of each line in the distribution network, determine the maximum node voltage and minimum node voltage in the voltage boundary constraints of the distribution network, and determine the maximum line current in the current boundary constraints of the distribution network; a first calculating module, used to calculate, for each node, a first difference between the maximum node voltage and the node voltage of the node, and calculate the ratio of the first difference to the rated voltage of the distribution network to obtain a first voltage per-unit value of the node; and a second calculating module, used to calculate a second difference between the node voltage and the minimum node voltage of the node, and calculate the second... The second voltage per-unit value of the node is obtained by comparing the difference with the rated voltage of the distribution network; the second determining module is used to determine the minimum voltage per-unit value from the first voltage per-unit value and the second voltage per-unit value of all nodes, and to determine the minimum voltage per-unit value as the first per-unit value; the third calculation module is used to calculate the third difference between the maximum line current and the line current of each line, and to calculate the ratio of the third difference to the maximum line current to obtain the current per-unit value of the line; the third determining module is used to determine the minimum current per-unit value from the current per-unit values ​​of all lines, and to determine the minimum current per-unit value as the second per-unit value.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for obtaining the power distribution network operation scheme as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method for obtaining the power distribution network operation scheme according to any one of claims 1 to 6.

Images