Power distribution network low-carbon high-quality scheduling method, system and equipment considering distribution robust optimization segmented affine and medium

Abstract

The invention discloses a power distribution network low-carbon high-quality scheduling method, system, equipment and medium considering distributed robust optimization segmented affine, and belongs to the technical field of power distribution network scheduling, and the method comprises the steps: taking the minimization of the carbon emission of the product of the main network carbon potential and the net load power as a target function; a Wasserstein distance is adopted to measure the deviation between historical data and real distribution, and multiple types of operation constraints are considered at the same time; the above distribution robust optimization model is converted and solved by adopting a piecewise affine theory, an original nonlinear optimization problem is converted into a linear optimization problem based on a simplex dominant uncertainty set, and the model solving efficiency and robustness are improved. According to the method, the uncertainty of new energy output and carbon potential is effectively dealt with through Wasserstein distance constraint, and the robustness is high; by means of segmented affine transformation, nonlinear robust optimization is simplified into linear optimization, and the solving speed is remarkably increased.

Description

Technical Field

[0001] This invention relates to the field of distribution network dispatching technology, specifically to a low-carbon, high-quality dispatching method, system, equipment, and medium for distribution networks that considers segmented affine optimization based on distributed bar. Background Technology

[0002] In modern power grid dispatching, balancing low-carbon operation with operational quality has become a crucial research topic. Existing technologies primarily address the uncertainties in renewable energy output and grid operation through stochastic programming or robust optimization methods. However, traditional stochastic programming methods typically rely on precise probability distribution information, which is often difficult to obtain accurately in practice, limiting the application of these methods. While traditional robust optimization methods can mitigate uncertainty to some extent, they are often overly conservative, leading to deviations from optimal dispatching results and reducing operational efficiency and low-carbon performance. Furthermore, practical applications suffer from high model complexity and low efficiency, making it difficult to meet the real-time and robustness requirements of power grids.

[0003] Therefore, existing methods suffer from shortcomings such as imperfect theory, insufficient robustness, and low computational efficiency when dealing with the randomness of new energy output and the dynamic uncertainty of the main grid carbon potential. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a method, system, equipment and medium for low-carbon and high-quality dispatching of distribution networks that considers segmented affine optimization of the split-bar system.

[0005] Therefore, the technical problem solved by this invention is: to construct an optimization model that can simultaneously consider the randomness of new energy output and the dynamic uncertainty of the main grid carbon potential, thereby ensuring the low-carbon nature of scheduling and operational quality; to improve the robustness and solution efficiency of the model while avoiding the over-conservatism of traditional robust optimization methods; and to realize the transformation of nonlinear robust optimization problems into linear optimization problems, thereby improving the solution speed while ensuring scheduling accuracy.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization with distributed blue bars, comprising, The objective function is to minimize the carbon emissions from the product of the main grid carbon potential and the net load power, where the net load power is the difference between the net load power before and after the integration of new energy sources and the implementation of regulation measures. The Wasserstein distance metric is used to measure the deviation between historical data and the actual distribution. A partial Bruker bar optimization model is constructed that includes the uncertainty of new energy output and the uncertainty of the main grid carbon potential. The operation constraints of the distribution network are established. The piecewise affine theory is used to transform the nonlinear robust optimization model. By using the equivalent substitution of uncertainty sets and taking advantage of the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, thus obtaining a distribution network dispatching decision scheme.

[0007] As a preferred embodiment of the low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on the split-loop rod, as described in this invention, the objective function of minimizing the carbon emissions of the product of the main grid carbon potential and the net load power includes decomposing the net load power into load regulation power and new energy power generation power. Carbon emission assessment indicators are obtained by summing the products of the main grid carbon potential and net load power of multiple nodes in the distribution network at multiple time periods. The objective function is constructed using a max-min operation, which seeks the optimal carbon emissions for scheduling decisions under uncertain scenarios.

[0008] As a preferred embodiment of the low-carbon and high-quality dispatching method for distribution networks considering segmented affine optimization of the sub-Bruker bar as described in this invention, the construction of the segmented Bruker bar optimization model, which includes uncertainty of new energy output and uncertainty of carbon potential of the main grid, includes constructing an uncertainty constraint set centered on the distribution of historical data samples using Wasserstein distance. The uncertainty constraint set is constrained by limiting the deviation range between the probability distribution of the sampled samples and the true probability distribution. We establish separate Wasserstein uncertainty constraints for new energy output and main grid carbon potential, and measure the probability distribution deviation of each by means of the joint probability distribution of the sampled samples. By combining the Wasserstein uncertainty constraint with the objective function of minimizing carbon emissions, a partial bar optimization model is constructed.

[0009] The beneficial effects of this preferred technical solution are as follows: by constructing an uncertainty constraint set through Wasserstein distance, when the probability distribution of new energy output and main grid carbon potential is difficult to obtain accurately, it can still effectively quantify the deviation between the historical data sample distribution and the true distribution, overcome the limitation of traditional stochastic programming relying on precise probability distribution, and at the same time avoid the problem of scheduling results deviating from the optimal due to excessive conservatism in traditional optimization methods.

[0010] As a preferred embodiment of the low-carbon and high-quality dispatching method for distribution networks considering piecewise affine optimization of the sub-Brubars as described in this invention, the step of transforming the sub-Brubars optimization model using piecewise affine theory includes constraining the sub-Brubars optimization model and using Wasserstein fuzzy set features to transform the sub-Brubars optimization model into an easily processed mathematical form. The original uncertainty set is equivalently replaced by the simplex-dominated uncertainty set, which is a convex polyhedron with multiple vertices; Identify the extreme locations of severe scenarios in the uncertainty set of the convex polyhedron, establish vertex constraints based on the extreme locations, and transform the nonlinear robust optimization problem into a linear optimization problem; Solve the linear optimization problem, derive the correspondence between the optimal solutions of the original uncertainty set and the uncertainty set after equivalent substitution, and obtain the power distribution network scheduling decision scheme.

[0011] The beneficial effects of this preferred technical solution are as follows: by replacing the original uncertainty set with the simplex-dominated uncertainty set, and by utilizing the mathematical characteristic that the adverse scenario is located at the pole of the convex polyhedron, the nonlinear optimization problem is transformed into a linear optimization problem, reducing the solution complexity from exponential to polynomial level. This significantly improves the model solution speed while ensuring scheduling accuracy, and meets the timeliness requirements of real-time scheduling of the distribution network.

[0012] As a preferred embodiment of the low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on the present invention, wherein: the establishment of distribution network operation constraints includes establishing power flow constraints, which include node power balance constraints, branch power transmission constraints and node voltage constraints; Establish network security constraints, which limit the node voltage range and line capacity upper limit by confidence levels; Establish active-reactive power management equipment constraints, which include reactive power compensation device constraints and on-load tap-changing transformer constraints.

[0013] As a preferred embodiment of the low-carbon and high-quality dispatching method for distribution networks considering segmented affine optimization of the blobs as described in this invention, wherein: the set of uncertainty constraints is a Wasserstein sphere, the Wasserstein sphere is defined by the infimum of the Wasserstein distance, and the infimum is the minimum value of the distance integral under the joint probability distribution of the sampled samples; The Wasserstein sphere is centered on the historical data sample distribution, and the deviation range of the probability distribution is limited by setting the radius of the Wasserstein sphere; The Wasserstein uncertainty constraints on the new energy output and the Wasserstein uncertainty constraints on the main grid carbon potential correspond to their respective Wasserstein sphere radii.

[0014] As a preferred embodiment of the low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on pluripotent bars as described in this invention, the objective function is expressed as follows: In the formula, Main grid carbon potential. The difference in net load power before and after the integration of new energy sources and the implementation of regulation measures. Carbon emissions reduced due to load changes Increased carbon contribution to new energy power generation.

[0015] This invention provides a low-carbon, high-quality dispatching system for distribution networks that considers segmented affine optimization based on distributed bar.

[0016] To solve the above technical problems, the present invention provides the following technical solution: a low-carbon and high-quality dispatching system for distribution networks considering segmented affine optimization of the distribution network, comprising: an objective function construction module, used to minimize the carbon emissions of the product of the main grid carbon potential and the net load power as the objective function, wherein the net load power is the difference between the net load power before and after the access of new energy sources and the adoption of regulation measures; The sub-Bruker bar optimization model construction module is used to construct a sub-Bruker bar optimization model that includes uncertainties in new energy output and carbon potential in the main grid by using the Wasserstein distance metric to measure the deviation between historical data and the actual distribution, and to establish operating constraints for the distribution network. The model transformation and solution module is used to transform the segmented robust optimization model using piecewise affine theory. By equivalent substitution of uncertainty sets and utilizing the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, thereby obtaining a distribution network dispatching decision scheme.

[0017] The present invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on distributed bar.

[0018] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the low-carbon and high-quality dispatching method for distribution networks considering segmented affine optimization of pluribars are implemented.

[0019] The beneficial effects of this invention are as follows: This invention uses the Wasserstein distance metric to measure the deviation between historical data and the actual distribution to construct a robust optimization model. This overcomes the limitations of traditional stochastic programming, which relies on precise probability distribution information, and avoids the problem of overly conservative scheduling results leading to deviations from the optimal solution. It can effectively quantify the range of uncertainty even when the probability distributions of renewable energy output and grid carbon potential are difficult to obtain accurately. By using piecewise affine theory to equivalently replace the original uncertainty set with the simplex-dominated uncertainty set, and utilizing the mathematical characteristics of severe scenarios located at the poles of convex polyhedra, the nonlinear optimization problem is transformed into a linear optimization problem. This significantly reduces the computational complexity of the model solution, improves the solution speed, and meets the timeliness requirements of real-time dispatching of the distribution network. Using the minimization of carbon emissions from the product of grid carbon potential and net load power as the objective function, the coordinated dispatching of active and reactive power management equipment maximizes the absorption of renewable energy. In high photovoltaic penetration scenarios, it solves the problems of node voltage exceeding limits and excessive backfeed power, achieving full grid connection of photovoltaic power generation without output reduction, and significantly improving the carbon emission reduction efficiency of the distribution network. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 The overall flowchart of a low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on a sub-blob bar is provided as an embodiment of the present invention.

[0022] Figure 2 The flowchart shows the piecewise affine solution process of a low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on a sub-Bruker bar, as provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0024] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a low-carbon, high-quality dispatching method for distribution networks that considers segmented affine optimization based on distributed bar optimization, including: Step 1: The objective function is to minimize the carbon emissions from the product of the main grid carbon potential and the net load power, where the net load power is the difference between the net load power before and after the new energy source is connected and regulation measures are taken. Step 2: Using Wasserstein distance to measure the deviation between historical data and the actual distribution, construct a sub-Bruker bar optimization model that includes uncertainties in new energy output and carbon potential in the main grid, and establish operating constraints for the distribution network. Step 3: The piecewise affine theory is used to transform the nonlinear robust optimization model. By using the equivalent substitution of the uncertainty set and taking advantage of the extreme point characteristics of the adverse scenario, the original nonlinear robust optimization problem is transformed into a linear optimization problem, and the power distribution network dispatch decision scheme is obtained.

[0025] In modern power grid dispatching, with the large-scale integration of new energy sources such as distributed photovoltaic (PV) power, the system faces the dual challenges of the randomness of new energy output and the dynamic uncertainty of the main grid's carbon potential. Traditional stochastic programming methods rely on precise probability distribution information of new energy output and main grid carbon potential, but in actual operation, these probability distributions are often difficult to obtain accurately, limiting the application of these methods. While traditional robust optimization methods can mitigate uncertainty to some extent, their overly conservative nature causes dispatching results to deviate from the optimal, reducing system operating efficiency and carbon sustainability. Furthermore, when PV penetration is high, the power grid is prone to safety issues such as node voltage exceeding limits and backfeed power exceeding line capacity. If effective dispatching measures are not taken, some PV grid-connected power must be reduced, resulting in a loss of carbon contribution. To address the aforementioned issues, this embodiment constructs an optimization model with the goal of minimizing carbon emissions in step 1, clarifying the low-carbon operation requirements of the scheduling; in step 2, it constructs a partial robust optimization model using Wasserstein distance, effectively quantifying uncertainty under conditions of incomplete probability distribution information, and simultaneously establishing operational constraints including power flow constraints, network security constraints, and active-reactive power management equipment constraints to ensure safe system operation; in step 3, it transforms the nonlinear robust optimization problem into a linear optimization problem using piecewise affine theory, significantly improving solution efficiency while ensuring scheduling accuracy.

[0026] The above three steps constitute a complete low-carbon, high-quality dispatching technology solution for distribution networks. Step 1 transforms the low-carbon operation requirements of the distribution network into a quantifiable mathematical optimization objective by establishing an objective function that is the product of the main grid carbon potential and net load power. Changes in net load power directly reflect the impact of renewable energy access and regulation methods on system carbon emissions. Step 2 utilizes a partial Bruker robust optimization model constructed using Wasserstein distance. This model establishes an uncertainty constraint set based on historical data sample distributions, even when the true probability distributions of renewable energy output and main grid carbon potential are unknown or difficult to obtain precisely. This avoids the dependence of traditional methods on precise distributions. Furthermore, the establishment of distribution network operation constraints ensures that the dispatching scheme meets both the low-carbon objective and the requirements for safe grid operation. Step 3 achieves a key breakthrough in model transformation through piecewise affine theory. It uses a simplex-dominated uncertainty set to equivalently replace the original uncertainty set and, based on the mathematical characteristic that adverse scenarios are located at the poles of convex polyhedra, transforms the originally complex nonlinear robust optimization problem into an easily solvable linear optimization problem. This allows the method to meet the computational speed requirements of real-time dispatching of distribution networks.

[0027] Example 2, refer to Figures 1-2 As one embodiment of the present invention, based on the previous embodiment, a low-carbon and high-quality dispatching method for distribution networks considering segmented affine optimization of distributed bars is provided, including: Step 1: Minimizing carbon emissions from the product of the main grid carbon potential and net load power is the objective function, which includes the following steps A1-A3: A1: The net load power is decomposed into load regulation power and new energy power generation power; A2: The carbon emission assessment index is obtained by summing the products of the main grid carbon potential and net load power of multiple nodes in the distribution network at multiple time periods; A3: The objective function is constructed using the max-min operation form, which seeks the optimal carbon emissions for scheduling decisions under uncertain scenarios.

[0028] In this embodiment of the application, in step 1, the objective function is to minimize the carbon emissions resulting from the product of the main grid carbon potential and the net load power. (1) (2) In the formula, Main grid carbon potential represents the amount of carbon dioxide emissions implied per unit power of electrical energy in the main grid. ), The difference in net load power before and after the integration of new energy sources and the implementation of regulation measures. Carbon emissions reduced due to load changes To improve the carbon contribution of new energy power generation. In sub-Bruker optimization, the max-min operation is usually used to represent the optimization of a certain objective function in the worst case. Specifically, in the uncertainty set, the difference between the maximum and minimum values ​​of the objective function under all possible uncertainty conditions is usually considered to find the worst case. max means to find the maximum objective value in all possible uncertainty scenarios. min means to minimize the maximum objective value, that is, to find the best decision in the worst case. In order to put the system in the optimal low-carbon operating state, the optimization objective function is set as the carbon potential of the main grid multiplied by the difference in net load power before and after the access of new energy and the implementation of adjustment measures, as shown in Equation (1); the net load can also be decomposed into the difference between the carbon emissions reduced by the load change and the carbon contribution improved by new energy power generation, as shown in Equation (2).

[0029] In an optional implementation, in step 1, minimizing the carbon emissions from the product of the grid carbon potential and net load power can be achieved by: first, decomposing the net load power according to its physical source. Changes in net load power mainly originate from two aspects: one is load-side power regulation, where demand response and other means cause changes in load power relative to a baseline value, directly affecting electricity purchased from the grid; the other is renewable energy generation, where distributed photovoltaic (PV) grid integration provides power within the distribution network, reducing the demand for electricity from the grid. When constructing the objective function, the contribution of these two parts to carbon emissions is calculated separately. For the load power change, the load regulation is obtained by subtracting the load power before and after regulation; multiplying this regulation by the grid carbon potential yields the carbon emission reduction from load regulation. For renewable energy generation, PV output directly contributes to reducing grid electricity purchases; multiplying this by the grid carbon potential yields the carbon contribution from renewable energy. Then, the total carbon emission change of the entire distribution network over the entire scheduling cycle is obtained by summing the results for all nodes and all scheduling periods of the distribution network. Finally, a two-layer optimization structure is adopted to represent the objective function: the outer layer optimization finds the optimal configuration of scheduling decision variables, and the inner layer optimization considers the uncertainty of the main grid carbon potential and the output of new energy sources, and finds the worst-case scenario among all possible uncertainty scenarios to ensure that the scheduling scheme can still maintain effective carbon emission reduction benefits under the worst conditions.

[0030] In another optional implementation, the objective function in step 1, which is to minimize the carbon emissions from the product of the grid carbon potential and net load power, can be achieved by: first, incorporating the impact of energy storage devices into the calculation of net load power. Energy storage devices have bidirectional power regulation capabilities; their charging process is equivalent to increasing the electrical load of the distribution network, thereby increasing the net load power; their discharging process is equivalent to providing power within the distribution network, thereby reducing the net load power. Therefore, when calculating the adjusted net load power, the combined effects of load power, energy storage charging power, energy storage discharging power, and photovoltaic power generation need to be comprehensively considered. Specifically, the adjusted net load power equals the load power plus the energy storage charging power, minus the energy storage discharging power and photovoltaic power generation. The difference between the net load power before and after adjustment is used to obtain the change in net load power caused by the dispatching measures. This change is multiplied by the grid carbon potential to obtain the carbon emission change brought about by the dispatching measures. Then, the carbon emission changes of all nodes in the distribution network during all dispatching periods are summed to obtain the overall carbon emission evaluation index. Finally, a two-layer optimization structure is adopted: the outer layer optimizes the configuration of scheduling decision variables such as the charging and discharging power of energy storage and photovoltaic output, while the inner layer seeks the most unfavorable scenario to address the uncertainties of the main grid's carbon potential and new energy output. In actual scheduling, energy storage can charge during peak photovoltaic output periods, absorbing excess photovoltaic power in the distribution network and avoiding backfeeding; it can discharge during peak load periods, reducing the purchase of electricity from the main grid, thereby minimizing carbon emissions.

[0031] Step 2: Using the Wasserstein distance metric to measure the deviation between historical data and the actual distribution, a partial Bruker bar optimization model is constructed, incorporating uncertainties in new energy output and main grid carbon potential. This establishes the distribution network operation constraints, including the following steps B1-B2: B1: A set of uncertainty constraints centered on the distribution of historical data samples is constructed using Wasserstein distance. The set of uncertainty constraints is constrained by limiting the range of deviation between the probability distribution of the sampled samples and the true probability distribution. B2: Establish separate Wasserstein uncertainty constraints for new energy output and main grid carbon potential, and measure the probability distribution deviation of each by means of the joint probability distribution of the sampled samples; B3: Combine the Wasserstein uncertainty constraint with the carbon emission minimization objective function to construct a partial Bruker bar optimization model.

[0032] In this embodiment of the application, in step B3, the Wasserstein uncertainty constraint is achieved through: (3) (4) (5) (6) In the formula, Let x be the probability distribution that the sampled sample follows; Let y be the probability distribution that the sampled sample follows; ε is the Wasserstein distance between probability distributions P and P1; ε is the radius of the Wasserstein sphere; D is the distance between probability distributions P and P1. A sphere with radius ε centered at ε; ; Let x be the joint probability distribution of samples x and y; This is the infimum. , For historical datasets, , The distribution is true. Based on the Wasserstein distance, the above two uncertainties are constructed to avoid the shortcomings of traditional stochastic programming's dependence on the exact probability distribution or the overly conservative nature of robust optimization. Among them, equation (5) represents the uncertainty of new energy output, and equation (6) represents the uncertainty of the main grid carbon potential.

[0033] In an optional implementation, in step B3, the Wasserstein uncertainty constraint can be achieved by: First, defining the Wasserstein distance as the minimum transmission cost between two probability distributions. Specifically, for the probability distribution P followed by the sampled samples and the surrogate probability distribution P1 of the true distribution, the Wasserstein distance is obtained by calculating the integral of the sample distances under all possible joint probability distributions and taking its infimum. The infimum represents the minimum value of the sample distance integral among all possible joint probability distributions, and this minimum value reflects the degree of true deviation between the two probability distributions. Then, an uncertainty constraint set is constructed using the Wasserstein distance. This set is defined as the set of all probability distributions whose Wasserstein distance from the historical data sample distribution does not exceed a given radius, forming a spherical region centered on the historical data sample distribution and bounded by a radius, called the Wasserstein sphere. For the uncertainty of new energy output, the Wasserstein distance is used to measure the deviation between the historical new energy output dataset and the true new energy output distribution, limiting the deviation to not exceed the radius of the Wasserstein sphere of new energy output. To address the uncertainty of the main grid carbon potential, the deviation between the historical main grid carbon potential dataset and the actual main grid carbon potential distribution is measured using the Wasserstein distance, with the deviation limited to the radius of the Wasserstein sphere of the main grid carbon potential. Finally, these two Wasserstein uncertainty constraints are incorporated into the optimization model and combined with the carbon emission minimization objective function to form a distributed blobs optimization model. This ensures that the optimization results remain effective for all possible probability distributions within the uncertainty range defined by the Wasserstein sphere.

[0034] In another optional implementation, in step B3, the Wasserstein uncertainty constraint can also be achieved by: First, establishing a sample set of renewable energy output and grid carbon potential based on historical operating data. For renewable energy output, collecting photovoltaic power generation data of each node in the distribution network during historical periods to form historical data samples of renewable energy output; for grid carbon potential, collecting carbon emission intensity data per unit power of electricity in the grid during historical periods to form historical data samples of grid carbon potential. Then, constructing an empirical probability distribution based on these historical data samples as an approximate proxy for the true probability distribution. The Wasserstein distance is used to measure the possible deviation between the empirical probability distribution and the true probability distribution, and the maximum allowable deviation range is controlled by setting the radius of the Wasserstein sphere. The setting of the Wasserstein sphere radius needs to comprehensively consider the number and quality of historical data samples. The more samples and the higher the data quality, the smaller the radius can be set; conversely, a larger radius needs to be set to cover a wider range of uncertainties. When constructing the sub-Brussels sphere optimization model, separate Wasserstein sphere constraints are established for both renewable energy output and main grid carbon potential. The distance between the sampled data and the true distribution is quantified by the integral form of the joint probability distribution, requiring that this distance does not exceed the respective set Wasserstein sphere radius. This constraint method enables the optimization model to cover the possible range of the true probability distribution even under conditions of incomplete historical data, thereby ensuring the effectiveness and reliability of the scheduling scheme in actual operation.

[0035] Furthermore, establishing distribution network operation constraints includes establishing power flow constraints, which include node power balance constraints, branch power transmission constraints, and node voltage constraints. Establish network security constraints, which limit the node voltage range and line capacity upper limit by confidence levels; Establish active-reactive power management equipment constraints, which include reactive power compensation device constraints and on-load tap-changing transformer constraints.

[0036] Specifically, the current flow constraint is represented as: (7) (8) (9) (10) (11) (12) In the formula, , These represent the active power injection and reactive power injection at node j at time t, respectively. ,、 They are power distribution lines The active power and reactive power flowing through branch ij at any given time; , These are power distribution lines. The active power and reactive power flowing out of node j and into node k at any given time; , They are power distribution lines The active and reactive power of the load at time node j; For power distribution lines The power of the reactive power compensation device installed at node j at time j; Let be the energy storage charging power of node j at time t in the power distribution line; Let be the energy storage discharge power of node j in the power distribution line at time t. The distributed power output of node j at time t in the power distribution line; for The current flows from node i to node j in branch at time t; branch road The resistance; branch road The reactance; For power distribution line nodes exist Voltage at a given moment.

[0037] Network security constraints are represented as: (13) (14) In the formula, This represents the confidence level at which the constraint holds. In the formula, Indicates the upper limit of line capacity. , This indicates the minimum and maximum allowable values ​​for node voltage.

[0038] Reactive power compensation constraints are expressed as: (15) (16) (17) In the formula, The set of nodes for installing capacitor banks; The number of operational groups represents the value of a discrete variable. This is the upper limit on the number of capacitor groups connected to node j; The compensation power for each group of capacitors. The set of nodes for installing static var compensators; and These are the lower and upper limits of the compensation power of the static reactive power compensation device, respectively. If it is a discrete reactive power compensation device, taking a capacitor bank as an example, the constraint is as shown in equation (15). If it is a continuous reactive power compensation device, taking a static reactive power compensation device as an example, the constraint is as shown in equation (17).

[0039] On-load tap changer constraint is represented as: (18) (19) In the formula, This is the set of substation nodes that include on-load tap-changing transformers; This is the high-voltage side voltage value of the transformer, which is a constant value. and It is the square of the upper and lower limits of the adjustable turns ratio of the on-load tap-changing transformer; The square of the turns ratio of an on-load tap-changing transformer is defined as the ratio of the secondary side to the primary side. It is a discrete variable and can be further processed into the following relationship containing 0-1 variables: (20) In the formula, This represents the difference between the square of the turns ratio of tap c and tap c-1 of an on-load tap-changing transformer, i.e., the adjacent adjustment increment. Variables are identified by 0-1.

[0040] The set of uncertainty constraints is a Wasserstein sphere, which is defined by the infimum of the Wasserstein distance, which is the minimum value of the distance integral under the joint probability distribution of the sampled samples. The Wasserstein sphere is centered on the historical data sample distribution, and the deviation range of the probability distribution is limited by setting the radius of the Wasserstein sphere; The Wasserstein uncertainty constraints on the new energy output and the Wasserstein uncertainty constraints on the main grid carbon potential correspond to their respective Wasserstein sphere radii.

[0041] Step 3: The piecewise affine theory is used to transform the nonlinear robust optimization model. Through equivalent substitution of the uncertainty set, and by utilizing the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, resulting in the distribution network dispatch decision scheme, which includes the following steps C1-C4: C1: Perform constraint optimization on the sub-Bruker bar optimization model, and use Wasserstein fuzzy set features to transform the sub-Bruker bar optimization model into an easily processed mathematical form; C2: The original uncertainty set is replaced by a simplex-dominated uncertainty set, which is a convex polyhedron with multiple vertices; C3: Identify the extreme positions of the adverse scenarios located in the uncertainty set of the convex polyhedron, establish vertex constraints based on the extreme positions, and transform the nonlinear robust optimization problem into a linear optimization problem; C4: Solve the linear optimization problem, derive the correspondence between the optimal solution of the original uncertainty set and the uncertainty set after equivalent substitution, and obtain the distribution network scheduling decision scheme.

[0042] In this embodiment of the application, in step 3, the piecewise affine theory is applied through: Model constraint optimization: (twenty one) In the formula, Represents the Wasserstein distance. This represents the optimal objective value of the original problem. This represents the optimal objective value of the equivalent problem. Let represent the radius of the Wasserstein sphere, which is used to define the size of the grazing level. This model is a convex optimization problem, and equation (21) utilizes the characteristics of Wasserstein fuzzy sets to transform it into an easily tractable mathematical form.

[0043] Equivalent substitution for uncertain sets: (twenty two) In the formula, This represents the scaling factor of the set. Let represent the i-th unit vector, and let represent the standard basis of the uncertainty space. This represents additional vertices, expanding the coverage of the set. Physically, it refers to a convex polyhedron with vertices at βe1,…, βem,βv, having a total of m+1 vertices. When m=3, the simplex-dominated uncertainty set replaces the 1-norm sphere uncertainty set; in this case, the adverse scenario in the max-min operation must lie within one of the four vertices.

[0044] Model transformation based on the set of uncertainties dominated by the simplex: (twenty three) In the formula, Let be the objective function. Let represent the scheduling decision variables. Equation (23) transforms the optimization model of this study into a compact form, with the aim of maximizing the solutions of all vertices. The minimum carbon benefit. Utilizing the extreme characteristics of severe scenarios located at the polyhedral uncertainty set, the above nonlinear robust optimization can be transformed into a linear optimization model: (twenty four) (25) In the formula, Indicates at the vertex The set of feasible solutions at a given location. Representing uncertain vertices, constraint (15) ensures that each xi is the corresponding uncertain vertex vi in The feasible solution is obtained by forcing the vertex position optimization solution to remain consistent without observing different uncertainties, while constraint (16) can effectively solve the problem of unpredictability restriction.

[0045] Model solution: (26) In the formula, , This represents the optimal solution of the original nonlinear sub-Bruker optimization problem expressed by equations (3)-(20); , This represents the optimal solution of the equivalent linear optimization problem expressed by equations (23)-(25). By using piecewise affine theory, the relationship between the optimal solution and the optimal value of the original uncertainty set and the equivalent replacement uncertainty set shown in equation (26) is derived, thus realizing the rapid solution of the original max-min model.

[0046] In an optional implementation, in step 3, piecewise affine theory can be applied as follows: First, the convex optimization properties of Wasserstein fuzzy sets are used to constrain the piecewise affine bar optimization model, transforming the original model into an equivalent form that is easier to solve. Then, a simplex-dominated uncertainty set is constructed to replace the original 1-norm sphere uncertainty set. The simplex set is a convex polyhedron structure with multiple vertices, and the position of each vertex is determined by the set scaling factor, unit vector, and additional vertices. Based on the mathematical property that adverse scenarios must be located at the vertices of this convex polyhedron, the original nonlinear optimization problem, which needs to be solved on the entire uncertainty set, is transformed into a linear optimization problem that only needs to be solved at a finite number of vertex positions. Finally, vertex constraints are established, requiring that the optimization solution at each vertex position maintains consistency across different uncertainty vertices. By solving the transformed linear optimization problem and deriving its correspondence with the optimal solution of the original problem, a distribution network scheduling decision scheme can be quickly obtained.

[0047] In another alternative implementation, step 3 can further employ piecewise affine theory by: first, equivalently replacing the original uncertainty set with a set of convex polyhedra dominated by the simplex. When the uncertainty space dimension is 3, this convex polyhedron contains 4 vertices, corresponding to 3 vertices in the unit vector directions and 1 additional vertex. The worst-case scenario is identified as occurring at one of these 4 vertices; therefore, it is unnecessary to consider all points within the uncertainty set, only the positions of these 4 vertices need optimization analysis. Then, corresponding feasibility constraints are established for each vertex to ensure that the scheduling decision satisfies all operational constraints under the uncertainty conditions corresponding to that vertex. By forcing consistency in scheduling decision variables at different vertices, the original model is transformed into a linear constraint optimization problem regarding vertex positions. Finally, utilizing the properties of piecewise affine functions, the linear relationship between the optimal solution under the original uncertainty set and the optimal solution under the simplex vertex set is derived, achieving an efficient linearized solution for the nonlinear robust optimization problem.

[0048] In this embodiment, step C4 solves the linear optimization problem by: establishing a linear optimization model based on vertex constraints; establishing corresponding feasible solution constraints for each uncertain vertex position; and ensuring that the scheduling decision satisfies the power flow constraints, network security constraints, and active-reactive power management device constraints of the distribution network under the uncertainty conditions corresponding to that vertex. By setting consistency constraints, the scheduling decision variables at all uncertain vertices are forced to remain consistent, that is, the scheduling decision must be effective for all possible vertex scenarios before the specific uncertainty realization value is observed. Based on the above vertex feasibility constraints and consistency constraints, a linear optimization problem is constructed and solved to obtain the optimal solution of the equivalent problem. Then, the correspondence between the optimal solution of the original uncertainty set and the equivalent simplex uncertainty set after substitution is derived using piecewise affine theory. Through this correspondence, the optimal solution of the equivalent problem is converted into the optimal solution of the original nonlinear partial bar optimization problem, thereby obtaining the scheduling decision scheme of the distribution network under uncertainty conditions, including the optimal configuration of decision variables such as photovoltaic output, energy storage charging and discharging power, on-load tap-changing transformer tap position, and the number of reactive power compensation device groups in operation at each node and time period.

[0049] In an optional implementation, in step C4, solving the linear optimization problem can be achieved by: first, transforming the original minimax optimization problem over the entire set of uncertainties into a minimum carbon benefit problem that maximizes all vertex solutions at a finite number of vertex positions. Specifically, for each uncertain vertex, calculate the carbon emission target value corresponding to the scheduling decision that satisfies all operational constraints under the conditions of that vertex. Then, take the minimum value among all vertex target values ​​as the carbon emission under the worst-case scenario. Finally, minimize the carbon emission under this worst-case scenario by optimizing the scheduling decision variables. This transformation process converts the nonlinear minimax problem into a standard linear programming problem. Then, use mature algorithms such as the simplex method or interior-point method of linear programming to solve the problem and obtain the optimal solution of the equivalent linear optimization problem. During the solution process, by introducing an auxiliary variable to represent the minimum value of all vertex target values ​​and adding corresponding linear inequality constraints, the objective function is transformed into a linear form that maximizes the auxiliary variable. Finally, based on the mapping relationship between the original problem and the optimal solution of the equivalent problem established by piecewise affine theory, the optimal solution of the equivalent problem is mapped back to the solution space of the original problem through linear transformation to obtain the optimal scheduling scheme of the original piecewise affine optimization problem.

[0050] In another optional implementation, in step C4, solving the linear optimization problem can also be achieved by: First, for each vertex of the simplex-dominated uncertainty set, establishing a set of linear constraints corresponding to that vertex position, including power flow balance constraints, voltage and current safety constraints, and equipment operation constraints under that vertex. Since the simplex set has a finite number of vertices, there are 4 vertices when the uncertainty space dimension is 3, so only 4 sets of linear constraints need to be established. Then, a consistency requirement is introduced into the optimization model, that is, the optimized scheduling decision must maintain the same value at the 4 different vertices. This requirement is achieved by adding equality constraints, making the scheduling decision adaptable to uncertainty. Based on the above constraints, a linear programming solver is used to solve the constructed linear optimization model to obtain the optimal scheduling decision under the vertex set. Finally, according to the piecewise affine function theory, the optimal solution of the original problem is equal to twice the optimal solution of the equivalent problem. The optimal solution of the original nonlinear robust optimization problem is directly calculated through this scaling relationship, achieving a fast solution.

[0051] Example 3, an embodiment of the present invention, provides a low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization with distributed bar. To verify the beneficial effects of the present invention, scientific demonstration is conducted through experiments: Taking a 10kV distribution line in the Donggang area of ​​Dalian as an example, the line has a uniformly distributed load and a total length of approximately 4.5km, mainly serving residential and commercial loads. With a high photovoltaic penetration rate (reaching 200%), problems such as node voltage exceeding the upper limit and excessive backfeed power have emerged. To study the regulation effect of the proposed method under different operating scenarios, several typical operating scenarios were constructed and compared.

[0052] In Scenario 1, assuming a large number of distributed photovoltaic (PV) systems are connected to both sides of the interconnected line, the node voltage exceeds 1.04 pu during peak PV output, causing some nodes to severely exceed their limits. Without intervention, some PV grid-connected power must be reduced, resulting in a loss of low-carbon contribution. Applying the proposed distributed bar optimization scheduling method combined with on-load tap regulation, the starting voltage was reduced from 1.04 pu to 1.02 pu. The results show that all node voltages returned to the allowable range, and no reduction in PV output was necessary, achieving full PV power absorption. Comparing the carbon contribution before and after, the system's carbon contribution increased from 729.2 kgCO2 to 812.6 kgCO2, an increase of approximately 11.4%, significantly enhancing the low-carbon operation effect.

[0053] In Scenario 2, considering the voltage over-limit issues on all interconnected lines, but with heavy loads on other outgoing lines and a positive power flow, simple on-load tap regulation might cause some lines to have voltages below 0.93 pu, affecting operational safety. Therefore, this method employs a strategy of coordinated control of inverter reactive power regulation and static reactive power compensation devices. After regulation, the overall node voltage decreased and returned to the normal range, while photovoltaic output achieved lossless grid connection. Results show that after installing SVC devices at nodes 6 and 16 and operating the inverters at a power factor of 0.9, the carbon contribution increased from 729.2 kgCO2 to 812.6 kgCO2, significantly improving the system's low-carbon level.

[0054] In Scenario 3, a large-scale photovoltaic (PV) power grid is connected to the end of the transmission line. During operation, both voltage limits are exceeded and backfeed power exceeds the line capacity. Traditional voltage regulation methods cannot meet these dual constraints in this scenario. This invention proposes introducing energy storage and load demand response into the scheduling to absorb excess power and transfer load demand during peak PV periods. In the example, the energy storage capacity is set at 50% of the peak PV output, and charging is performed at maximum power during peak PV output. The results show that the integration of energy storage effectively reduces the voltage level at the end of the line and the backfeed load, ensuring that all PV power is connected to the grid for consumption. The system's carbon contribution increases from 729.2 kgCO2 to 1003.4 kgCO2, an increase of over 37%, fully demonstrating the significant role of energy storage in improving low-carbon efficiency. Simultaneously, the demand response mechanism increases load power by 30% during midday, further alleviating voltage and backfeed issues and providing support for safe system operation.

[0055] Finally, in complex scenarios, when photovoltaic penetration further increases to 250% and multiple nodes simultaneously experience voltage and power limit exceedances, single adjustment methods are insufficient to guarantee the safety and low-carbon operation of the system. Using the proposed sub-blob rod optimization and piecewise affine solution method, optimized scheduling is performed under multiple uncertainties. The results show that by prioritizing the absorption of surplus power by energy storage, combined with on-load tap changer to reduce bus voltage levels, and supplemented by inverter reactive power regulation and static reactive power compensation, the system ultimately achieves comprehensive satisfaction of voltage and power flow constraints, maximizing the grid-connected absorption of photovoltaic power generation. These results verify the adaptability and coordination of this method under complex operating conditions, significantly improving the renewable energy absorption capacity and the system's low-carbon operation level.

[0056] In summary, the example cases clearly demonstrate the scheduling effectiveness of the proposed method in different typical scenarios. By introducing bibru bar optimization under Wasserstein distance and piecewise affine transformation, this invention effectively handles the uncertainty of photovoltaic output and the dynamic changes in grid carbon potential while maintaining computational efficiency, comprehensively considering voltage security constraints, power constraints, and low-carbon goals.

[0057] Example 4 is an embodiment of the present invention, which provides a low-carbon, high-quality dispatching system for distribution networks that considers segmented affine optimization based on distributed bar, including: The objective function construction module is used to minimize the carbon emissions of the product of the main grid carbon potential and the net load power as the objective function, wherein the net load power is the difference between the net load power before and after the access of new energy sources and the implementation of regulation measures; The sub-Bruker bar optimization model construction module is used to construct a sub-Bruker bar optimization model that includes uncertainties in new energy output and carbon potential in the main grid by using the Wasserstein distance metric to measure the deviation between historical data and the actual distribution, and to establish operating constraints for the distribution network. The model transformation and solution module is used to transform the segmented robust optimization model using piecewise affine theory. By equivalent substitution of uncertainty sets and utilizing the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, thereby obtaining a distribution network dispatching decision scheme.

[0058] This embodiment also provides an electronic device applicable to the low-carbon and high-quality dispatching method for distribution networks considering segmented affine transformation with sparse bar optimization, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the low-carbon and high-quality dispatching method for distribution networks considering segmented affine transformation with sparse bar optimization as proposed in the above embodiment.

[0059] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the low-carbon, high-quality dispatching method for distribution networks that considers segmented affine optimization based on distributed bar optimization as proposed in the above embodiments.

[0060] The storage medium proposed in this embodiment and the low-carbon, high-quality dispatching method for distribution networks that considers segmented affine optimization based on distributed bar optimization proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0061] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0062] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization with distributed bar optimization, characterized by: include, The objective function is to minimize the carbon emissions from the product of the main grid carbon potential and the net load power, where the net load power is the difference between the net load power before and after the integration of new energy sources and the implementation of regulation measures. The Wasserstein distance metric is used to measure the deviation between historical data and the actual distribution. A partial Bruker bar optimization model is constructed, which includes the uncertainty of new energy output and the uncertainty of the main grid carbon potential. The operation constraints of the distribution network are established. The piecewise affine theory is used to transform the original nonlinear robust optimization model. By using the equivalent substitution of uncertainty sets and taking advantage of the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, thus obtaining a distribution network dispatching decision scheme.

2. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar optimization as described in claim 1, characterized in that: The objective function of minimizing carbon emissions from the product of the main grid carbon potential and net load power includes decomposing the net load power into load regulation power and new energy power generation power. Carbon emission assessment indicators are obtained by summing the products of the main grid carbon potential and net load power of multiple nodes in the distribution network at multiple time periods. The objective function is constructed using a max-min operation, which seeks the optimal carbon emissions for scheduling decisions under uncertain scenarios.

3. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar as described in claim 2, characterized in that: The construction of the sub-Bruker bar optimization model, which includes uncertainty of new energy output and uncertainty of main grid carbon potential, includes constructing an uncertainty constraint set centered on the distribution of historical data samples using Wasserstein distance. The uncertainty constraint set is constrained by limiting the deviation range between the probability distribution of the sampled samples and the true probability distribution. We establish separate Wasserstein uncertainty constraints for new energy output and main grid carbon potential, and measure the probability distribution deviation of each by means of the joint probability distribution of the sampled samples. By combining the Wasserstein uncertainty constraint with the objective function of minimizing carbon emissions, a partial bar optimization model is constructed.

4. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar as described in claim 3, characterized in that: The transformation of the sub-Brubar optimization model using piecewise affine theory includes constraining the sub-Brubar optimization model and using Wasserstein fuzzy set features to transform the sub-Brubar optimization model into an easily processed mathematical form. The original uncertainty set is equivalently replaced by the simplex-dominated uncertainty set, which is a convex polyhedron with multiple vertices; Identify the extreme locations of severe scenarios in the uncertainty set of the convex polyhedron, establish vertex constraints based on the extreme locations, and transform the nonlinear robust optimization problem into a linear optimization problem; Solve the linear optimization problem, derive the correspondence between the optimal solutions of the original uncertainty set and the uncertainty set after equivalent substitution, and obtain the power distribution network scheduling decision scheme.

5. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar as described in claim 4, characterized in that: The establishment of distribution network operation constraints includes establishing power flow constraints, which include node power balance constraints, branch power transmission constraints, and node voltage constraints. Establish network security constraints, which limit the node voltage range and line capacity upper limit by confidence levels; Establish active-reactive power management equipment constraints, which include reactive power compensation device constraints and on-load tap-changing transformer constraints.

6. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar as described in claim 5, characterized in that: The set of uncertainty constraints is a Wasserstein sphere, which is defined by the infimum of the Wasserstein distance, which is the minimum value of the distance integral under the joint probability distribution of the sampled samples. The Wasserstein sphere is centered on the distribution of historical data samples, and the deviation range of the probability distribution is limited by setting the radius of the Wasserstein sphere; The Wasserstein uncertainty constraints on the new energy output and the Wasserstein uncertainty constraints on the main grid carbon potential correspond to their respective Wasserstein sphere radii.

7. The low-carbon, high-quality dispatching method for distribution networks considering segmented affine optimization based on distributed bar optimization as described in claim 6, characterized in that: The objective function is expressed as follows: In the formula, Main grid carbon potential. The difference in net load power before and after the integration of new energy sources and the implementation of regulation measures. Carbon emissions reduced due to load changes Increased carbon contribution to new energy power generation.

8. A low-carbon, high-quality dispatching system for distribution networks considering segmented affine transformations with sub-Bruker optimization, employing the low-carbon, high-quality dispatching method for distribution networks considering segmented affine transformations with sub-Bruker optimization as described in any one of claims 1 to 7, characterized in that... include: The objective function construction module is used to minimize the carbon emissions of the product of the main grid carbon potential and the net load power as the objective function, wherein the net load power is the difference between the net load power before and after the access of new energy sources and the implementation of regulation measures; The Wasserstein distance metric is used to measure the deviation between historical data and the actual distribution, construct a Wasserstein distance metric to build a Wasserstein distance metric that includes uncertainties in new energy output and carbon potential in the main grid, and establish operating constraints for the distribution network. The model transformation and solution module is used to transform the segmented robust optimization model using piecewise affine theory. By equivalent substitution of uncertainty sets and utilizing the extreme point characteristics of severe scenarios, the original nonlinear robust optimization problem is transformed into a linear optimization problem, thereby obtaining a distribution network dispatching decision scheme.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on split-bars, as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the low-carbon and high-quality dispatching method for distribution networks that considers segmented affine optimization based on split-bars, as described in any one of claims 1 to 7.

Images