A power distribution grid double-layer collaborative optimization control method based on capability constraint transmission, a storage medium and a system

Abstract

This invention discloses a two-layer collaborative optimization control method, storage medium, and system for distribution grids based on capability constraint propagation, belonging to the field of distribution network optimization control technology. The method includes: an offline phase, where the lower-level controller undergoes scanning tests within its boundary power range, and a capability constraint function is established through piecewise linear fitting; and an online phase, where the upper-level optimizer uses the capability constraint function as an explicit constraint to optimize the allocation of photovoltaic power reduction at each node, and sends the results down to the lower-level controller for minute-level voltage collaborative control. This invention, through a capability constraint propagation mechanism, quantifies the uncertain capability boundary of the lower-level controller offline into deterministic constraints, eliminating execution failures caused by capability mismatch in two-layer collaboration, effectively solving the voltage limit exceedance problem in high-PV-penetration distribution networks, and improving the robustness and reliability of system operation.

Description

Technical Field

[0001] This invention belongs to the field of power distribution network optimization control technology, specifically relating to a two-layer collaborative optimization control method, storage medium, and system for power distribution grids based on capability constraint transmission. Background Technology

[0002] With the large-scale integration of distributed power sources such as photovoltaics into distribution networks, the strong fluctuations and intermittent nature of photovoltaic power output can easily lead to voltage exceeding the upper limit at local nodes of the distribution network during periods of sufficient sunlight, threatening the safe and stable operation of the system. To address this issue, the two-level collaborative control method, which divides optimization decision-making into hourly-level upper-level scheduling and minute-level lower-level control, can better adapt to the multi-timescale fluctuation characteristics of photovoltaic power output and has become the mainstream research direction.

[0003] However, existing two-layer cooperative control methods mainly have the following problems:

[0004] 1) When formulating photovoltaic reduction and allocation schemes, upper-level optimization usually assumes that the lower-level controller has full execution capabilities and does not quantify and constrain the actual adjustment capabilities of the lower level. When the upper-level command exceeds the lower-level execution capability, it will lead to voltage over-limit or control failure.

[0005] 2) Some studies solve the upper and lower layer models simultaneously, but the scale of the simultaneous models is huge and the solution complexity is high, making it difficult to meet the requirements of online real-time performance;

[0006] 3) The method of using iterative coordination for information exchange between upper and lower layers is difficult to guarantee convergence and has a heavy communication burden;

[0007] 4) The capability boundary of the lower-level controller is affected by multiple factors such as network topology, load level and equipment capacity, and is therefore uncertain. It is difficult to describe directly in the upper-level optimization control model, resulting in insufficient robustness of the two-level collaboration.

[0008] There is currently no effective solution to the above problems. Summary of the Invention

[0009] The purpose of this invention is to provide a two-layer collaborative optimization control method, storage medium, and system for distribution grids based on capability constraint transfer. By quantifying the uncertain capability boundary of the lower-level controller into deterministic constraints, the invention avoids execution failures caused by capability mismatch of the lower-level controller during two-layer collaborative control, thereby improving the robustness and reliability of the system operation. The technical solution adopted by this invention is as follows.

[0010] On one hand, the present invention provides a two-layer collaborative optimization control method for distribution grids based on capability constraint transfer, comprising:

[0011] Based on historical operation data of the power distribution network, typical scenarios corresponding to various levels of excess photovoltaic power are generated.

[0012] For each typical scenario, select boundary power limit test points according to the set boundary power scan range and step size;

[0013] In each typical scenario, the lower-level optimization control model is solved based on the test points of each boundary power limit to obtain the photovoltaic reduction feasibility state and photovoltaic reduction amount. The capacity boundary of each typical scenario is obtained according to the change of photovoltaic reduction amount, and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted and used as the lower-level capacity constraint function.

[0014] Based on the real-time operating status data of the distribution network, the typical scenario is matched, and the lower-level capacity constraint function corresponding to the matched typical scenario is used as one of the constraints of the upper-level optimization control model. The upper-level optimization control model is solved to obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and then sent to the lower-level controller to perform voltage coordinated control.

[0015] It should be noted that the photovoltaic reduction feasibility state obtained from the lower-level optimization control model, i.e., at the selected boundary power limit test point, is whether there exists a feasible solution that simultaneously satisfies the node voltage constraint and the branch power flow constraint.

[0016] Optionally, the various photovoltaic excess power levels corresponding to the typical scenario include: a high photovoltaic excess power level, corresponding to a photovoltaic excess power range of [missing information]. The level of excess photovoltaic power in China, and the corresponding range of excess photovoltaic power. Low photovoltaic excess power level, corresponding to the photovoltaic excess power range is: .

[0017] Optionally, the solution of the lower-level optimization control model based on the test points of each boundary power limit includes: solving a mixed integer linear programming model with the objective of minimizing the total photovoltaic reduction of the distribution network, while satisfying the node voltage constraints and branch power flow constraints;

[0018] The objective function of the mixed-integer linear programming model is expressed as: ,in, For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network.

[0019] The correlation between nodal photovoltaic power reduction and distribution network operating status parameters is expressed as follows:

[0020] ,

[0021] ,

[0022] in, and The first Actual injected power and predicted photovoltaic output of each photovoltaic node For nodes The load power, For nodes The net injected power of a node directly affects the node voltage level, constitutes the input state quantity of the lower-level voltage collaborative control, and is transmitted to the node voltage constraint through the branch power flow equation, forming a complete coupling relationship between photovoltaic power reduction and node operating state.

[0023] The node voltage constraint is expressed as: ,in, For the first The voltage amplitude at each node, , These are the lower and upper limits of the node voltage, respectively;

[0024] Branch flow constraints are represented as: ,in, For branch circuits in the distribution network The meritorious trend, This is the upper limit of the thermally stable transmission capacity of the branch;

[0025] The relationship between branch active power flow and node voltage is expressed as follows:

[0026] ,

[0027] in, , They are nodes , voltage amplitude, For nodes , Voltage phase angle difference between them branch road The impedance.

[0028] Optionally, when solving the lower-level optimization control model based on each boundary power limit test point in each typical scenario, the lower-level optimization control model is solved sequentially based on each boundary power limit test point in ascending order. This will reduce the total photovoltaic reduction of the distribution network to 0, and the photovoltaic reduction feasibility status of the lower-level optimization control model will be feasible. The boundary power limit test points will be used as the capability boundary in the current typical scenario.

[0029] The lower-level capacity constraint function is a piecewise function, obtained by fitting the linear relationship between the power limits on both sides of the capacity boundary and the total photovoltaic reduction of the distribution network.

[0030] Optionally, the lower-level capability constraint function is expressed as:

[0031] ,

[0032] in, Indicated by boundary power limit The lower-level capacity constraint function is the independent variable, i.e., the relationship function between the boundary power limit and the total photovoltaic reduction of the distribution network; The fitted slope, The intercept is... The capability boundary is the maximum boundary power limit that the lower-level controller can maintain node voltage compliance without reducing photovoltaic power.

[0033] Optionally, the method of the present invention further includes, after fitting the lower-level capability constraint function, calculating the goodness of fit. To evaluate its fitting accuracy, if If the accuracy falls below the preset accuracy threshold, the density of boundary power limit test points is increased, the lower-level optimization control model is resolved, and a new lower-level capability constraint function is obtained through refitting, until the refitted lower-level capability constraint function is found to be within acceptable limits. If the accuracy requirements are met, then it is determined as the final lower-level capability constraint function;

[0034] Among them, goodness of fit The calculation formula is:

[0035] ,

[0036] in, For the first The actual reduction at each test point These are the fitted values. This represents the average reduction amount across all test points.

[0037] Optionally, the upper-level optimization control model aims to minimize the total photovoltaic reduction in the distribution network, and solves for the optimal photovoltaic reduction allocation scheme and the corresponding photovoltaic power range for each photovoltaic node in the distribution network; the constraints of the upper-level optimization control model include capacity constraint function constraints, network topology constraints, and power balance constraints.

[0038] The objective function of the upper-level optimization control model is expressed as: , For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network.

[0039] Capability constraint function constraints are expressed as follows: , For the first The boundary power limit of each photovoltaic node, and its corresponding capacity constraint function. The function value is the first Photovoltaic reduction per photovoltaic node.

[0040] It should be noted that although the objective function forms of the lower-level optimization control model and the upper-level optimization control model are the same, they differ in several aspects, including the optimization period, decision content, and decision input information. For specific differences, please refer to existing two-level cooperative control technologies.

[0041] Optionally, the step of matching the typical scenario based on the real-time operating status data of the distribution network, and using the lower-level capacity constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model, and solving the upper-level optimization control model, includes:

[0042] Based on the real-time collected photovoltaic output and load data of each node, calculate the total excess photovoltaic power of the distribution network at the current moment. :

[0043] in, and Representing photovoltaic nodes Photovoltaic power output and load data;

[0044] Will Match the excess photovoltaic power range corresponding to the typical scenario, select the lower-level capacity constraint function corresponding to the matching typical scenario as one of the constraints of the upper-level optimization control model, and solve the upper-level optimization control model.

[0045] For example, when When selecting a high-excess scenario, choose the capability constraint function corresponding to that scenario. ;when When selecting the capability constraint function corresponding to the moderate excess scenario, choose the appropriate one. ;when When selecting a scenario with low excess capacity, choose the capability constraint function corresponding to that scenario. .

[0046] In the above technical solution, the capacity constraint function matched by the real-time state of the distribution network is embedded as an explicit upper bound constraint into the upper-level optimization control model. Together with the network topology constraint and power balance constraint, it forms a constraint set. By calling the mixed integer linear programming solver, the optimal reduction allocation scheme and corresponding boundary power limit of each photovoltaic node in the regional distribution network can be obtained. This is then sent down to the lower-level controller to execute the corresponding voltage coordination control, realizing the online activation and complete closure of the coupling constraint relationship between the upper and lower layers.

[0047] In a second aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the two-layer collaborative optimization control method for power distribution grids based on capability constraint transmission as described in the first aspect.

[0048] Thirdly, the present invention provides a two-layer collaborative optimization control system for power distribution grid based on capability constraint transmission, which includes an upper-layer controller and a lower-layer controller.

[0049] The upper-layer controller includes:

[0050] The capacity constraint function construction unit is configured to generate typical scenarios corresponding to various photovoltaic excess power levels based on historical operating data of the distribution network during the offline phase; for each typical scenario, boundary power limit test points are selected according to the set boundary power scanning range and step size; under each typical scenario, the lower-level optimization control model is solved based on each boundary power limit test point to obtain the photovoltaic reduction amount and feasibility state; the capacity boundary of each typical scenario is obtained based on the change of photovoltaic reduction amount; and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted to obtain the linear reduction interval and zero reduction interval of photovoltaic reduction amount, which are used as the lower-level capacity constraint function.

[0051] The collaborative control unit is configured to, during the online operation phase, match the typical scenario based on the real-time operating status data of the distribution network, and use the lower-level capability constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model to solve the upper-level optimization control model, obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and send it down to the lower-level controller.

[0052] The lower-level controller: According to the optimal photovoltaic reduction scheme, within the boundary power range constraint, it performs minute-level voltage coordinated control, and maintains the voltage of each node within the qualified range by adjusting the reactive power output and active power reduction of the photovoltaic inverter at each photovoltaic node.

[0053] Beneficial effects

[0054] Compared with the prior art, the present invention has the following advantages and advancements:

[0055] 1. This invention proposes a capability constraint transfer mechanism for a two-layer collaborative optimization control method for distribution grids based on capability constraint transfer. By using offline scanning tests, the uncertain capability boundary of the lower-level controller is quantified into a deterministic piecewise linear constraint function, which is then embedded into the upper-level optimization control model in the form of explicit constraints. This fundamentally eliminates the execution failure problem caused by capability mismatch in the two-layer collaboration and effectively solves the problem of voltage limit exceeding in distribution networks with high photovoltaic penetration.

[0056] 2. This invention transforms the lower-level capability boundary into explicit constraints that can be directly invoked by the upper level through offline pre-computation. This allows the upper-level online optimization to complete the decision without repeatedly iterating and calling the lower-level model in each control cycle, which significantly reduces the computational burden of online solution, meets the timeliness requirements of real-time control of distribution networks, and has good engineering practicality.

[0057] 3. This invention quantifies the accuracy of the capability constraint function by using the goodness-of-fit R², and uses R² not lower than a preset threshold as the condition for judging the effectiveness of the fit, which ensures the accuracy of capability constraint transmission and provides a solid guarantee for the robustness and reliability of the two-layer collaborative control.

[0058] 4. This invention treats the lower-level controller as a black box for capability boundary testing, without requiring modification or disclosure of its internal model structure. It has good versatility and compatibility, is applicable to various types of lower-level controllers, and is easy to promote and apply in engineering.

[0059] 5. This invention uses piecewise linear functions to describe capability constraints, and the resulting constraints are linear expressions that can be directly embedded into the upper-level mixed integer linear programming model without changing its linear programming properties. This ensures the solvability and solution efficiency of the upper-level model and is suitable for large-scale power distribution network engineering scenarios. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 The diagram shows a flowchart of the two-layer collaborative optimization control method for power distribution grid based on capability constraint transmission according to the present invention.

[0062] Figure 2 The diagram shows the principle of the two-layer collaborative optimization control based on capability constraint transmission of the present invention.

[0063] Figure 3 The figure shown is a comparison of the test data of capability constraints in various scenarios and the piecewise linear fitting results of this invention.

[0064] Figure 4 The figure shown is a performance evaluation chart of the capability constraint fitting for various scenarios of the present invention. Detailed Implementation

[0065] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] Example 1

[0067] This embodiment introduces a two-layer collaborative optimization control method for distribution grids based on capability constraint transfer, which includes:

[0068] Based on historical operation data of the power distribution network, typical scenarios corresponding to various levels of excess photovoltaic power are generated.

[0069] For each typical scenario, select boundary power limit test points according to the set boundary power scan range and step size;

[0070] In each typical scenario, the lower-level optimization control model is solved based on the test points of each boundary power limit to obtain the photovoltaic reduction feasibility state and photovoltaic reduction amount. The capacity boundary of each typical scenario is obtained according to the change of photovoltaic reduction amount, and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted and used as the lower-level capacity constraint function.

[0071] Based on the real-time operating status data of the distribution network, the typical scenario is matched, and the lower-level capacity constraint function corresponding to the matched typical scenario is used as one of the constraints of the upper-level optimization control model. The upper-level optimization control model is solved to obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and then sent to the lower-level controller to perform voltage coordinated control.

[0072] refer to Figure 1 The specific implementation of this embodiment involves the following parts.

[0073] 1. Generate a photovoltaic overcapacity test scenario and set the boundary power scan range and step size parameters.

[0074] Specifically, photovoltaic (PV) output in the distribution network is affected by weather conditions, exhibiting significantly different excess power levels at different times. To comprehensively cover various operating conditions in actual operation, some implementation methods construct several typical scenarios based on historical distribution network operating data and PV excess power levels. These typical scenarios at least cover: high PV excess power levels, corresponding to a PV excess power range of [missing information]. The level of excess photovoltaic power in China, and the corresponding range of excess photovoltaic power. Low photovoltaic excess power level, corresponding to the photovoltaic excess power range is: .

[0075] For each typical scenario, the starting value, ending value, and step size of the boundary power range are set. The boundary power range is then discretized and scanned to obtain multiple boundary power limit test points. .

[0076] In this embodiment, the IEEE 33-node distribution network is used as the test system to construct three typical photovoltaic (PV) overcapacity scenarios: S1, a high overcapacity scenario with a PV overcapacity of 257kW, representing a high PV overcapacity level; S2, a low overcapacity scenario with a PV overcapacity of 71.25kW, representing a low PV overcapacity level; and S3, a medium overcapacity scenario with a PV overcapacity of 145.55kW, representing a medium PV overcapacity level. For each scenario, the starting value for the boundary power limit scan is set to 50kW, with step sizes of S1-20kW, S2-5kW, and S3-10kW, respectively.

[0077] By covering multiple scenarios, the capability constraint function constructed subsequently can accurately reflect the actual execution capability of the lower-level controller under different photovoltaic output levels of the distribution network, ensuring that the coupling relationship between the upper and lower-level optimization control models is effectively described under various operating conditions.

[0078] 2. Solve the lower-level optimization model point by point and record the photovoltaic reduction amount and feasibility status at each test point.

[0079] Specifically, after determining a series of boundary power limit test points corresponding to each typical scenario, based on each boundary power limit test point, with the goal of minimizing the total photovoltaic reduction in the distribution network, and under the conditions of satisfying node voltage constraints and branch power flow constraints, the lower-level optimization control model is solved to obtain the photovoltaic reduction feasibility state and photovoltaic reduction amount.

[0080] This invention treats the lower-level controller as a black box and systematically probes the response of the lower-level controller to each possible boundary power decision of the upper-level controller, thereby quantifying the coupling constraint relationship between the upper and lower layers.

[0081] In some embodiments, the upper-level optimization control model is a mixed-integer linear programming model, and its objective function is expressed as: ,in, For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network.

[0082] The correlation between nodal photovoltaic power reduction and distribution network operating status parameters is expressed as follows:

[0083] ,

[0084] ,

[0085] in, and The first Actual injected power and predicted photovoltaic output of each photovoltaic node For nodes The load power, For nodes The net injected power at a node directly affects the node voltage level, constitutes the input state quantity for lower-level voltage collaborative control, and is transmitted to the node voltage constraint through the branch power flow equation, forming a complete coupling relationship between photovoltaic power reduction and node operating state.

[0086] The node voltage constraint is expressed as: ,in, For the first The voltage amplitude at each node, , These are the lower and upper limits of the node voltage, respectively;

[0087] Branch flow constraints are represented as: ,in, For branch circuits in the distribution network The meritorious trend, This is the upper limit of the thermally stable transmission capacity of the branch;

[0088] The relationship between branch active power flow and node voltage is expressed as follows:

[0089] ,

[0090] in, , They are nodes , voltage amplitude, For nodes , Voltage phase angle difference between them branch road The impedance.

[0091] The aforementioned photovoltaic (PV) reduction feasibility state refers to whether, under the current boundary power limit, the lower-level mixed-integer linear programming model has a feasible solution for the optimal PV reduction amount that simultaneously satisfies the node voltage constraint and the branch power flow constraint. The state value is either "feasible" or "infeasible".

[0092] Through the lower-level optimization control model, the feasibility status of photovoltaic (PV) reduction and the corresponding total PV reduction in the distribution network can be collected at test points with different boundary power limits under various typical scenarios. For example... Figure 3 As shown, under various typical scenarios, with the boundary power limit... The increase in the total amount of photovoltaic power reduction across the entire network The reduction trend is linear, until it reaches zero. At each boundary power limit test point, if the model's photovoltaic reduction feasibility is deemed infeasible, the corresponding boundary power limit test point is not included in the fitted data. As the boundary power limit test points... The step size is gradually increased until the total photovoltaic power reduction of the entire network corresponding to a certain test point is reached. When the value drops to 0 and the model remains feasible, this test point can serve as the capability boundary. This refers to the maximum boundary power limit that the lower-level controller can maintain without reducing photovoltaic power, which is the maximum boundary power limit that allows the node voltage to be within acceptable limits.

[0093] III. Constructing capability constraint functions through piecewise linear fitting.

[0094] In some possible embodiments, when solving the lower-level optimization control model based on each boundary power limit test point in each typical scenario, the lower-level optimization control model is solved in order from smallest to largest based on each boundary power limit test point. This will reduce the total photovoltaic reduction of the distribution network to 0, and the photovoltaic reduction feasibility status of the lower-level optimization control model will be feasible. The boundary power limit test points will be used as the capability boundary in the current typical scenario.

[0095] The lower-level capacity constraint function is a piecewise function, obtained by fitting the linear relationship between the power limits on both sides of the capacity boundary and the total photovoltaic reduction of the distribution network.

[0096] The lower-level capability constraint function is expressed as:

[0097] ,

[0098] in, This represents the lower-level capacity constraint function, i.e., the relationship between the boundary power limit and the total photovoltaic reduction in the distribution network; The fitted slope, The intercept is... The capability boundary is the maximum boundary power limit that the lower-level controller can maintain node voltage compliance without reducing photovoltaic power.

[0099] In some possible embodiments, after fitting the lower-level capability constraint function, the goodness of fit is also calculated. To evaluate its fitting accuracy, if If the accuracy falls below the preset accuracy threshold, the density of boundary power limit test points is increased, the lower-level optimization control model is resolved, and a new lower-level capability constraint function is obtained through refitting, until the refitted lower-level capability constraint function is found to be within acceptable limits. If the accuracy requirements are met, then it is determined as the final lower-level capability constraint function;

[0100] Among them, goodness of fit The calculation formula is:

[0101] ,

[0102] in, For the first The actual reduction at each test point These are the fitted values. This represents the average reduction amount across all test points.

[0103] refer to Figure 3 and Figure 4 In this embodiment, the test results for three typical scenarios are as follows: Scenario S1 meets the boundary power limit test point. During the process of increasing from 50kW to 250kW The capacity decreases linearly from 207.0 kW to near 0, therefore the capacity boundary... S2 scenario in During the process of increasing from 50kW to 70kW The capacity decreases linearly from 21.2 kW to near 0, therefore the capacity boundary... S3 scene in During the process of increasing from 50kW to 140kW The power decreases linearly from 95.6 kW to near 0, therefore the capacity boundary... The fitting and validation results of the lower-level capability constraint functions for three typical scenarios are shown in Table 1 below.

[0104] Table 1

[0105]

[0106] Table 1 shows the fitting slopes for the three scenarios. All are -1.0, intercept Each of these parameters closely matches the excess photovoltaic power levels in various typical scenarios, with clear physical meaning; capacity boundaries The values ​​are 250 kW, 70 kW, and 140 kW, respectively, accurately reflecting the maximum execution capability of the lower-level controller under different excess levels. The R² accuracy of the three typical scenarios is no less than 0.999, indicating that the piecewise linear fitting has high accuracy and the constructed capability constraint function can accurately describe the execution capability boundary of the lower-level controller, providing a reliable guarantee for the explicit transmission of the coupling constraint relationship between the upper and lower levels.

[0107] The content of parts one through three above can all be implemented by the host controller during the offline phase.

[0108] IV. Online Phase: Acquire real-time operating status data of the distribution network.

[0109] During the online operation and control phase, real-time operational status data, including load power, photovoltaic output, and network topology status at each node of the distribution network, are collected. This data serves as input to the upper-level optimization control model, providing a basis for subsequent optimization decisions based on the current system state. The collected real-time operational status data, together with the capability constraint function constructed in the offline phase, constitutes the complete input to the upper-level optimization control model, reflecting the real-time activation of the upper-level coupling relationship during the online phase.

[0110] Fifth, embed the lower-level capacity constraint function into the upper-level optimization model to optimize the decision-making of photovoltaic reduction schemes for each node.

[0111] Since corresponding lower-level capability constraint functions are constructed for different typical scenarios during the offline phase, the first step is to match typical scenarios based on the real-time operating status of the distribution network in order to select appropriate ones. .

[0112] In this embodiment, the total excess photovoltaic power of the distribution network at the current moment is calculated based on the real-time collected photovoltaic output and load data of each node. :

[0113] in, and Representing photovoltaic nodes Photovoltaic power output and load data;

[0114] Will Match the excess photovoltaic power range corresponding to the typical scenario, select the lower-level capacity constraint function corresponding to the matching typical scenario as one of the constraints of the upper-level optimization control model, and solve the upper-level optimization control model.

[0115] For example, when When selecting a high-excess scenario, choose the capability constraint function corresponding to that scenario. ;when When selecting the capability constraint function corresponding to the moderate excess scenario, choose the appropriate one. ;when When selecting a scenario with low excess capacity, choose the capability constraint function corresponding to that scenario. .

[0116] At this point, the capacity constraint function matched by the real-time state of the distribution network is embedded as an explicit upper bound constraint into the upper-level optimization control model. Together with the network topology constraint and power balance constraint, it forms a constraint set. By calling the mixed-integer linear programming solver, the optimal reduction allocation scheme and the corresponding boundary power limit of each photovoltaic node in the regional distribution network can be obtained.

[0117] In this embodiment, the upper-level optimization control model aims to minimize the total photovoltaic (PV) power reduction in the distribution network, solving for the optimal PV power reduction allocation scheme and corresponding PV power range for each PV node in the distribution network. The constraints of the upper-level optimization control model include capacity constraint functions, network topology constraints, and power balance constraints. That is, the aforementioned offline-constructed lower-level capacity constraint function... By embedding explicit upper bound constraints on the photovoltaic reduction of each node into the upper-level optimization model, the upper-level decision-making is always constrained by the lower-level execution capability, thus realizing the online activation of the coupling relationship between the upper and lower levels.

[0118] The objective function of the upper-level optimization control model is expressed as: , For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network.

[0119] Capability constraint function constraints are expressed as follows: .

[0120] Although the objective function forms of the lower-level optimization control model and the upper-level optimization control model are the same, they differ in several aspects, including the optimization cycle, decision content, and decision input information. Specifically:

[0121] 1. In terms of time scale: The upper-level optimization control model is coarse-grained, at the hour level, and makes decisions on the planned photovoltaic reduction allocation scheme within a relatively long time window; the lower-level optimization control model is fine-grained, at the minute level, and performs optimization adjustments at the real-time execution level.

[0122] 2. Different decision-making content: The upper-level decision is "how much photovoltaic reduction is allocated to each photovoltaic node", which determines the boundary power limit of each photovoltaic node; the lower-level decision is "how to adjust the reactive power output and active power reduction of the inverter under given boundary constraints so that the voltage of each node meets the constraints", which is a specific voltage control execution problem.

[0123] 3. Differences in information completeness: The input to the upper-level decision-making is the global photovoltaic power output forecast and load forecast, and it cannot obtain the detailed model structure of the lower-level internals; the input to the lower-level decision-making is the local accurate electrical parameters and real-time status, and it receives the boundary constraints issued by the upper level as input.

[0124] 6. Send the optimization results to the lower-level controller for minute-level voltage collaborative control.

[0125] The boundary power limits of each node obtained from the upper-level optimization control model are sent to the lower-level controller as constraint boundary conditions. Within the received boundary constraints, the lower-level controller performs minute-level voltage coordinated control, maintaining the voltage of each node within the acceptable range by adjusting the reactive power output and active power reduction of each photovoltaic inverter. Since the upper-level optimization has ensured that the issued commands are within the lower-level execution capability range by embedding capability constraint functions, the coupling relationship between the upper and lower levels is completely closed, and the lower-level controller can guarantee successful execution without voltage exceeding the limit or control failure.

[0126] Example 2

[0127] Based on the same inventive concept as Embodiment 1, this embodiment introduces a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the power distribution grid two-layer collaborative optimization control method based on capability constraint transmission as described in Embodiment 1.

[0128] Example 3

[0129] Based on the same inventive concept as Embodiments 1 and 2, this embodiment introduces a two-layer collaborative optimization control system for power distribution grids based on capability constraint transfer, referencing... Figure 2 As shown, it includes an upper-level controller and a lower-level controller;

[0130] The upper-layer controller includes:

[0131] The capacity constraint function construction unit is configured to generate typical scenarios corresponding to various photovoltaic excess power levels based on historical operating data of the distribution network during the offline phase; for each typical scenario, boundary power limit test points are selected according to the set boundary power scanning range and step size; under each typical scenario, the lower-level optimization control model is solved based on each boundary power limit test point to obtain the photovoltaic reduction amount and feasibility state; the capacity boundary of each typical scenario is obtained based on the change of photovoltaic reduction amount; and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted to obtain the linear reduction interval and zero reduction interval of photovoltaic reduction amount, which are used as the lower-level capacity constraint function.

[0132] The collaborative control unit is configured to, during the online operation phase, match the typical scenario based on the real-time operating status data of the distribution network, and use the lower-level capability constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model to solve the upper-level optimization control model, obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and send it down to the lower-level controller.

[0133] The lower-level controller: According to the optimal photovoltaic reduction scheme, within the boundary power range constraint, it performs minute-level voltage coordinated control, and maintains the voltage of each node within the qualified range by adjusting the reactive power output and active power reduction of the photovoltaic inverter at each photovoltaic node.

[0134] Figure 2 In this context, the capability constraint function construction unit includes:

[0135] The scenario production module is used to generate typical scenarios corresponding to various photovoltaic excess power levels based on historical operation data of the distribution network during the offline phase.

[0136] The capability testing module is used to select boundary power limit test points for each typical scenario according to the set boundary power scanning range and step size; under each typical scenario, the lower-level optimization control model is solved based on each boundary power limit test point to obtain the photovoltaic reduction amount and feasibility status.

[0137] The fitting module is used to obtain the capability boundaries of each typical scenario based on the changes in photovoltaic reduction, and to fit the linear reduction interval and zero reduction interval of photovoltaic reduction, which are used as the lower-level capability constraint function.

[0138] The collaborative control unit includes:

[0139] The data acquisition module is used to collect real-time operating status data of the power distribution network during the online operation phase;

[0140] The upper-level optimization control module is used to match the typical scenario based on the real-time operating status data of the distribution network, and use the lower-level capacity constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model to solve the upper-level optimization control model, obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and send it down to the lower-level controller.

[0141] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention 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.

[0142] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. 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 illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0143] 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.

[0144] 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.

[0145] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A power distribution grid bi-level collaborative optimization control method based on capability constraint transmission, characterized in that, include: Based on historical operation data of the power distribution network, typical scenarios corresponding to various levels of excess photovoltaic power are generated. For each typical scenario, select boundary power limit test points according to the set boundary power scan range and step size; In each typical scenario, the lower-level optimization control model is solved based on the test points of each boundary power limit to obtain the photovoltaic reduction feasibility state and photovoltaic reduction amount. The capacity boundary of each typical scenario is obtained according to the change of photovoltaic reduction amount, and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted and used as the lower-level capacity constraint function. Based on the real-time operating status data of the distribution network, the typical scenario is matched, and the lower-level capacity constraint function corresponding to the matched typical scenario is used as one of the constraints of the upper-level optimization control model. The upper-level optimization control model is solved to obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and then sent to the lower-level controller to perform voltage coordinated control.

2. The method according to claim 1, characterized in that, The typical scenarios correspond to various levels of excess photovoltaic power, including: high excess photovoltaic power level, with a corresponding range of excess photovoltaic power. The level of excess photovoltaic power in China, and the corresponding range of excess photovoltaic power. Low photovoltaic excess power level, corresponding to the photovoltaic excess power range is: .

3. The method according to claim 1, characterized in that, The solution of the lower-level optimization control model based on the test points of each boundary power limit includes: solving a mixed integer linear programming model with the objective of minimizing the total photovoltaic reduction of the distribution network, while satisfying the node voltage constraints and branch power flow constraints; The objective function of the mixed-integer linear programming model is expressed as: ,in, For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network. The correlation between nodal photovoltaic power reduction and distribution network operating status parameters is expressed as follows: ， ， in, and The first Actual injected power and predicted photovoltaic output of each photovoltaic node For nodes The load power, For nodes Net injection power; The node voltage constraint is expressed as: ,in, For the first The voltage amplitude at each node, , These are the lower and upper limits of the node voltage, respectively. Branch flow constraints are represented as: ,in, For branch circuits in the distribution network The meritorious trend, This is the upper limit of the thermally stable transmission capacity of the branch; The relationship between branch active power flow and node voltage is expressed as follows: ， in, , They are nodes , voltage amplitude, For nodes , Voltage phase angle difference between them branch road The impedance.

4. The method according to claim 1, characterized in that, When solving the lower-level optimization control model based on each boundary power limit test point in each typical scenario, the lower-level optimization control model is solved in order from smallest to largest based on each boundary power limit test point. This will reduce the total photovoltaic reduction of the distribution network to 0, and the photovoltaic reduction feasibility status of the lower-level optimization control model will be feasible. The boundary power limit test points are used as the capability boundary in the current typical scenario. The lower-level capacity constraint function is a piecewise function, obtained by fitting the linear relationship between the power limits on both sides of the capacity boundary and the total photovoltaic reduction of the distribution network.

5. The method according to claim 1, characterized in that, The lower-level capability constraint function is expressed as follows: ， in, Indicated by boundary power limit Let be the lower-level capability constraint function of the independent variable. The fitted slope, The intercept is... The capability boundary is the maximum boundary power limit that the lower-level controller can maintain node voltage compliance without reducing photovoltaic power.

6. The method according to claim 1, characterized in that, This also includes calculating the goodness of fit after fitting the lower-level capability constraint function. To evaluate its fitting accuracy, if If the accuracy falls below the preset accuracy threshold, the density of boundary power limit test points is increased, the lower-level optimization control model is resolved, and a new lower-level capability constraint function is obtained through refitting, until the refitted lower-level capability constraint function is found to be within acceptable limits. If the accuracy requirements are met, then it is determined as the final lower-level capability constraint function; Among them, goodness of fit The calculation formula is: ， in, For the first The actual reduction at each test point These are the fitted values. This represents the average reduction amount across all test points.

7. The method according to claim 6, characterized in that, The upper-level optimization control model aims to minimize the total photovoltaic (PV) reduction in the distribution network, and solves for the optimal PV reduction allocation scheme and corresponding PV power range for each PV node in the distribution network. The constraints of the upper-level optimization control model include capacity constraint function constraints, network topology constraints, and power balance constraints. The objective function of the upper-level optimization control model is expressed as: , For the first Power reduction per photovoltaic node This represents the total number of photovoltaic nodes in the distribution network. Capability constraint function constraints are expressed as follows: , For the first The boundary power limit of each photovoltaic node, and its corresponding capacity constraint function. The function value is the first Photovoltaic reduction per photovoltaic node.

8. The method according to claim 7, characterized in that, The process of matching typical scenarios based on real-time operating status data of the distribution network, and using the lower-level capacity constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model, and solving the upper-level optimization control model includes: Based on the real-time collected photovoltaic output and load data of each node, calculate the total excess photovoltaic power of the distribution network at the current moment. : ， in, and Representing photovoltaic nodes Photovoltaic power output and load data; Will Match the excess photovoltaic power range corresponding to the typical scenario, select the lower-level capacity constraint function corresponding to the matching typical scenario as one of the constraints of the upper-level optimization control model, and solve the upper-level optimization control model.

9. 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 two-layer collaborative optimization control method for power distribution grid based on capability constraint transmission as described in any one of claims 1 to 8.

10. A two-layer collaborative optimization control system for power distribution grids based on capability constraint transfer, characterized in that, Includes upper-level controllers and lower-level controllers; The upper-layer controller includes: The capacity constraint function construction unit is configured to generate typical scenarios corresponding to various photovoltaic excess power levels based on historical operating data of the distribution network during the offline phase; for each typical scenario, boundary power limit test points are selected according to the set boundary power scanning range and step size; under each typical scenario, the lower-level optimization control model is solved based on each boundary power limit test point to obtain the photovoltaic reduction amount and feasibility state; the capacity boundary of each typical scenario is obtained based on the change of photovoltaic reduction amount; and the linear reduction interval and zero reduction interval of photovoltaic reduction amount are fitted to obtain the linear reduction interval and zero reduction interval of photovoltaic reduction amount, which are used as the lower-level capacity constraint function. The collaborative control unit is configured to, during the online operation phase, match the typical scenario based on the real-time operating status data of the distribution network, and use the lower-level capability constraint function corresponding to the matched typical scenario as one of the constraints of the upper-level optimization control model to solve the upper-level optimization control model, obtain the optimal photovoltaic reduction allocation scheme and corresponding boundary power limit for each photovoltaic node in the entire network, and send it down to the lower-level controller. The lower-level controller: According to the optimal photovoltaic reduction scheme, within the boundary power range constraint, it performs minute-level voltage coordinated control, and maintains the voltage of each node within the qualified range by adjusting the reactive power output and active power reduction of the photovoltaic inverter at each photovoltaic node.

Images