A power transmission and distribution collaborative economic dispatch method considering high proportion of distributed energy access

Abstract

The application discloses a kind of transmission and distribution collaborative economic dispatching method under considering high proportion distributed energy access, belong to the technical field of power system automation, the method is aimed at strong uncertainty and short time volatility under high proportion distributed energy access, and a kind of day-ahead-intra-day multi-time scale transmission and distribution collaborative optimization framework is proposed.In day-ahead stage, a two-stage robust optimization model is constructed to solve the "worst" scenario of wind and solar load to ensure system robustness, and the transmission and distribution distributed collaborative optimization is carried out through target cascade method to develop global day-ahead scheduling plan;In intra-day stage, to correct the conservative bias of day-ahead and cope with short-term severe fluctuations, the rolling optimization idea of model predictive control is introduced, and the transmission and distribution distributed collaborative optimization is carried out again through target cascade method in 15 minutes scale, the application effectively considers the robustness and economy of system operation, realizes the privacy protection and multi-time scale collaborative operation of transmission and distribution network.

Description

Technical Field

[0001] This invention belongs to the field of power system automation and economic dispatch technology, specifically relating to a transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access. Background Technology

[0002] With the continued advancement of the "carbon peaking and carbon neutrality" goals, a large proportion of distributed renewable energy sources such as wind power and photovoltaics are being integrated into the distribution network, transforming it from a traditional one-way receiving-end network into an active distribution network with bidirectional power flow. However, the power output and load demand of wind and solar power are highly uncertain and volatile, posing a dual challenge to the safe and economical operation of the system.

[0003] On the one hand, in the face of system uncertainty, traditional methods often adopt single-stage robust optimization, which is based on the characteristic of finding the best solution based on extreme boundaries. This often leads to overly conservative scheduling results and high backup costs. On the other hand, traditional independent optimization of transmission and distribution networks has severed the mutual support capability of flexible resources, while centralized optimization faces the dual barriers of data privacy and computational dimension.

[0004] On the other hand, renewable energy exhibits significant short-term fluctuations. Relying solely on day-ahead scheduling plans on a single timescale makes it difficult to capture drastic short-term fluctuations, leading to power deficits in actual implementation. If energy storage devices were to handle all real-time power compensation, frequent and disorderly start-ups and shutdowns would severely impact their lifespan. Therefore, there is an urgent need for a transmission and distribution co-optimization method that can overcome uncertainty and protect privacy while adapting to short-term fluctuations and preserving energy storage lifespan. Summary of the Invention

[0005] To solve the above-mentioned technical problems, or at least partially solve them, this invention proposes a transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a method for coordinated economic dispatch of transmission and distribution considering a high proportion of distributed energy access, comprising:

[0007] S1: Day-ahead uncertainty handling stage: Based on the day-ahead wind and solar load forecast data, a two-stage robust optimization TS-RO model containing the box uncertainty set of wind and solar load is constructed with a time scale of 1 hour; with the goal of minimizing the system operating cost when the uncertain variables take the worst operating scenario in the uncertainty set, the TS-RO model is decomposed into a main problem and sub-problems in mixed integer linear form and solved to obtain the worst operating scenario of wind and solar load;

[0008] S2: Day-ahead global scheduling phase: Taking the worst-case operating scenario obtained in S1 as a deterministic input, and aiming to minimize the overall operating cost of the transmission and distribution network, a transmission and distribution coordinated optimization scheduling model based on the objective cascade method (ATC) is constructed. The transmission network and distribution network are decoupled through boundary interaction power. The transmission network treats the interaction power as a virtual load, and the distribution network treats the interaction power as a virtual generator. After the transmission and distribution networks are solved independently, only the boundary power information is coordinated to obtain the day-ahead optimal scheduling strategy for transmission and distribution coordination.

[0009] S3: Intraday Rolling Correction Phase: Introducing the rolling optimization concept of Model Predictive Control (MPC), with a 15-minute time scale and a 4-hour prediction time domain, the start-up and shutdown status of thermal power units and the basic operation plan of energy storage in the day-ahead optimal scheduling strategy obtained in S2 are used as deterministic constraints; using real-time updated ultra-short-term wind and solar load forecast data, an intraday phase multi-time scale rolling collaborative optimization model based on ATC is constructed, and a double penalty cost for energy storage deviating from the day-ahead plan is introduced into the intraday objective function; the optimal transmission and distribution coordination plan for the next 4 hours is solved through the ATC algorithm, and only the scheduling instructions for the first 15-minute period are executed. Then, the prediction time domain is shifted forward by one period, and the above process is repeated until the scheduling of all 96 periods of the day is completed.

[0010] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in this invention, the day-ahead TS-RO dispatch model specifically includes the construction of the wind and solar load box uncertainty set:

[0011] To describe the uncertainty of wind and solar load, a box-shaped uncertainty set for the fluctuation range of wind and solar load is constructed, specifically represented as follows:

[0012]

[0013] Where U is the box-type uncertain set of wind and solar loads. For a set of uncertain variables, , , Distribution network During the period The values ​​of wind power output, photovoltaic power output, and load active power under uncertain scenarios, where B is a set of binary variables. , , These are 0-1 variables representing whether the uncertainties of wind power, solar power, and load have reached their boundaries. When the value is 1, the wind power and solar power output reaches their minimum values, and the load power reaches its maximum values ​​for the corresponding time period. This represents the maximum fluctuation deviation of wind power output in the distribution network. This represents the predicted wind power output of the distribution network. This represents the maximum fluctuation deviation of photovoltaic power output in the distribution network. This is the predicted value of photovoltaic output in the distribution network. This represents the maximum fluctuation deviation of the active power of the distribution network load. This represents the predicted value of the active power of the distribution network load. , , These are the uncertainty adjustment parameters for wind and solar loads introduced into the distribution network, with values ​​ranging from 0 to 24 (integers). For distribution network numbering, For the dispatch period, DSO represents the distribution network side.

[0014] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access described in this invention, the solution process of the day-ahead TS-RO dispatch model is as follows:

[0015] By adjusting the uncertainty adjustment parameter The conservative nature of the scheduling scheme is flexibly adjusted; the min problem in the inner layer of the TS-RO model is transformed into a max problem using strong duality theory, and after merging with the max problem in the outer layer, the original model is decoupled into a main problem and sub-problems in mixed integer linear form; the column and constraint generation algorithm (C&CG) is used to iteratively solve the main problem and sub-problems until the convergence condition is met, thus obtaining the worst-case operating scenario.

[0016] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method under the high proportion of distributed energy access described in this invention, the transmission and distribution coordinated decoupling mechanism based on ATC in S2 is specifically: the transmission and distribution coordinated optimization dispatch model based on ATC in the day-ahead stage, which realizes the decoupling and coordination between the transmission network and the distribution network through boundary interactive power;

[0017] The interactive power is equivalent to virtual generators on the distribution network side and virtual loads on the transmission network side, and consistency constraints are introduced:

[0018]

[0019] in, for The transmission network transmits to the first time period The interactive power of the distribution network for Time period The interactive power transmitted from the distribution network to the transmission network.

[0020] By adding penalty terms based on the Lagrange multiplier method to the independent objective functions of the transmission and distribution networks, the reconstructed day-ahead ATC objective function is as follows:

[0021]

[0022]

[0023] in, The total operating cost of the power transmission network, For the first The total operating cost of a power distribution network For the number of distribution networks, and For the multiplier of the penalty term in the ATC algorithm, and The interaction power value is obtained from the adjacent region. The superscript "-" indicates that this item is a known quantity.

[0024] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method under the high proportion of distributed energy access described in this invention, the total operating cost of the transmission and distribution network in S2 includes: the power generation cost, start-up and shutdown cost, reserve cost, renewable energy power generation cost and transmission and distribution interaction cost of the thermal power units on the transmission network side; and the operating cost of energy storage devices, interruptible load compensation cost, transferable load adjustment cost, charging and discharging cost of electric vehicles using V2G mode, renewable energy power generation cost and transmission and distribution interaction cost on the distribution network side.

[0025] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access described in this invention, the solution process for the intraday rolling coordinated optimization in S3 is as follows:

[0026] In the current prediction time domain Within the current forecast, 15-minute ultra-short-term wind and solar load forecast data is obtained. Based on the current forecast data, the intraday ATC model is iteratively solved until the interactive power of the transmission and distribution network meets the convergence condition. The multi-period coordinated scheduling results within the current forecast domain are output. The optimal scheduling command for the first 15-minute period is extracted and issued for execution. Then, t=t+1 is set, and the forecast time domain is slid forward by one period. The above optimization process is repeated until the scheduling of all 96 periods of the day is completed.

[0027] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access described in this invention, wherein: the expression for the double penalty cost of energy storage deviating from the day-ahead plan in S3 is:

[0028]

[0029]

[0030] in, This is the intraday period. For the intraday rolling optimization of the prediction time domain, and The penalty cost for energy storage devices during the daytime phase of the distribution network, and The penalty coefficient is... and For the intraday phase of the distribution network Energy storage charging and discharging status during a given period of time. and This represents the day-ahead charging and discharging status of energy storage in the distribution network. and For the intraday phase of the distribution network Energy storage charging and discharging power during a given period and This represents the day-ahead energy storage charging and discharging power of the distribution network.

[0031] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access described in this invention, the day-ahead ATC model of S2 and the intraday ATC model of S3 adopt a unified convergence criterion and multiplier update rule, and the convergence condition is:

[0032]

[0033] in, and The convergence coefficient is . For the number of iterations, For the region No. The objective function value of the next iteration. For the region No. The objective function value of the next iteration.

[0034] If the convergence condition is not met, then update the penalty function factor:

[0035]

[0036] in, It is a constant. and The initial value is a relatively small constant. and The first The algorithm's penalty term multiplier in the next iteration, and These are the multipliers for the penalty term in the updated algorithm. For the first Second iteration The transmission network transmits to the first time period The interactive power of the distribution network For the first iteration Time period The interactive power transmitted from the distribution network to the transmission network.

[0037] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access described in this invention, wherein: in S2, the distribution network forms an adjustable capacity boundary through joint dispatch of various flexible resources within the network, specifically:

[0038] The power distribution network coordinates the scheduling of energy storage devices, interruptible loads, transferable loads, and electric vehicles using the V2G mode to calculate the upper limit of adjustable capacity and the lower limit of adjustable capacity. By constraining the power transmission and distribution interaction within the physically feasible range through the adjustable capacity boundary, the power interaction command that exceeds the limit is avoided.

[0039] As a preferred embodiment of the transmission and distribution coordinated economic dispatch method considering high-proportion distributed energy access described in this invention, when converting the TS-RO model into a mixed-integer linear form, the product terms of binary variables and continuous variables in the model are linearized using the Big M method, wherein the Big M coefficients corresponding to the auxiliary variables range from 10³ to 10⁻⁶. 6 .

[0040] The beneficial effects of this invention are: 1. Enhanced operational robustness and risk resistance: By constructing a box-type uncertainty set for wind and solar loads and using TS-RO to solve the "worst-case" scenario, the uncertainty risk brought about by the high proportion of distributed energy access is effectively resisted.

[0041] 2. Balancing global economic efficiency and data privacy: Using transmission and distribution interaction power as a coupling variable, distributed collaborative optimization is carried out using ATC, which achieves the optimal total cost of day-ahead global dispatch while ensuring the information privacy of each power grid entity.

[0042] 3. Improved scheduling adaptability and real-time accuracy: By embedding ATC into the rolling loop of MPC and combining it with real-time high-precision data, the day-ahead plan is dynamically corrected on a 15-minute time scale, which effectively overcomes the day-ahead forecast error and further improves the actual operating economy. Attached Figure Description

[0043] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0044] Figure 1 A schematic diagram of the interconnected system structure of a 6-node transmission network and 2-node distribution network provided by the present invention;

[0045] Figure 2 This is a schematic diagram of the day-to-day scheduling framework for the coordinated transportation and distribution provided by the present invention;

[0046] Figure 3A schematic diagram of the solution process for the intraday phase ATC-based multi-timescale optimization scheduling model for transport and distribution coordination is provided in this embodiment of the invention.

[0047] Figure 4 This is a schematic diagram of the intraday power balance and equipment output of the power transmission and distribution network provided in an embodiment of the present invention;

[0048] Figure 5 This is a schematic diagram of the power balance and equipment output of the power distribution network within one day, provided by an embodiment of the present invention.

[0049] Figure 6 This is a schematic diagram of the power balance and equipment output of the power distribution network within 2 days, provided as an embodiment of the present invention. Detailed Implementation

[0050] To make the above-mentioned objects, features and advantages of 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.

[0051] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0052] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0053] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0054] Example 1

[0055] Reference Figures 1-3 The first embodiment of the present invention provides a method for coordinated economic dispatch of transmission and distribution considering a high proportion of distributed energy access, comprising the following steps:

[0056] S1: Day-ahead uncertainty handling stage: Based on the day-ahead wind and solar load forecast data, a two-stage robust optimization TS-RO model containing the box uncertainty set of wind and solar load is constructed with a time scale of 1 hour; with the goal of minimizing the system operating cost when the uncertain variables take the worst operating scenario in the uncertainty set, the TS-RO model is decomposed into a main problem and sub-problems of mixed integer linear form and solved to obtain the worst operating scenario of wind and solar load.

[0057] First, we construct a multi-timescale optimization scheduling model for coordinated transportation and distribution:

[0058] (1) Day-ahead scheduling model for coordinated transmission and distribution

[0059] The objective function of the transmission network is to minimize the global operating cost of the transmission network, including the generation cost of generating units, start-up and shutdown costs, reserve costs, renewable energy generation costs, and interaction costs.

[0060] The objective function of the distribution network is to minimize the global operating cost of the distribution network, including the operating cost of energy storage devices, the cost of demand response scheduling, the cost of charging and discharging electric vehicles, the cost of renewable energy generation, and the interaction cost.

[0061] Power transmission network constraints: physical operation constraints of generating units, ramping constraints, start-up and shutdown constraints, and wind and solar power output constraints;

[0062] Distribution network constraints: energy storage device operation constraints, demand response operation constraints, electric vehicle operation constraints, and wind and solar power output constraints;

[0063] Optimization target: Optimize the optimal scheduling plan of the power transmission and distribution network within 24 hours of the following day, using 1 hour as the time scale.

[0064] (2) Intraday rolling phase model of transmission and distribution coordination

[0065] The objective function of the power transmission network is to minimize the actual daily operating cost.

[0066] Power grid constraints: Inherit the start-up and shutdown status constraints of thermal power units in the day-ahead stage, and update the high-precision ultra-short-term wind and solar load forecast boundary;

[0067] The objective function for the distribution network is to minimize the sum of the actual daily operating costs and the penalty costs for deviating from the daily plan.

[0068] The intraday objective function of the distribution network k incorporates a penalty cost f to prevent frequent fluctuations in energy storage:

[0069]

[0070]

[0071] in, This is the intraday period. For the intraday rolling optimization of the prediction time domain, and The penalty cost for energy storage devices during the daytime phase of the distribution network, and The penalty coefficient is... and For the intraday phase of the distribution network Energy storage charging and discharging status during a given period of time. and This represents the day-ahead charging and discharging status of energy storage in the distribution network. and For the intraday phase of the distribution network Energy storage charging and discharging power during a given period and This represents the day-ahead energy storage charging and discharging power of the distribution network.

[0072] Distribution network constraints: Inherit the day-ahead energy storage device scheduling plan and update the high-precision ultra-short-term wind and solar load forecast boundary;

[0073] Optimization target: Using 15 minutes as the time scale and 4 hours as the prediction time domain, the rolling optimization outputs refined scheduling instructions and transmission and distribution interaction power of various flexible resources (demand response, electric vehicles, etc.) in the next 4 hours.

[0074] TS-RO algorithm solution process:

[0075] First, to describe the uncertainty of wind and solar load, a box-shaped uncertainty set is constructed to represent the fluctuation range of wind and solar load, specifically as follows:

[0076]

[0077] Where U is the box-type uncertain set of wind and solar loads. For a set of uncertain variables, , , Distribution network During the period The values ​​of wind power output, photovoltaic power output, and load active power under uncertain scenarios, where B is a set of binary variables. , , These are 0-1 variables representing whether the uncertainties of wind power, solar power, and load have reached their boundaries. When the value is 1, the wind power and solar power output reaches their minimum values, and the load power reaches its maximum values ​​for the corresponding time period. This represents the maximum fluctuation deviation of wind power output in the distribution network. This represents the predicted wind power output of the distribution network. This represents the maximum fluctuation deviation of photovoltaic power output in the distribution network. This is the predicted value of photovoltaic output in the distribution network. This represents the maximum fluctuation deviation of the active power of the distribution network load. This represents the predicted value of the active power of the distribution network load. , , These are the uncertainty adjustment parameters for wind and solar loads introduced into the distribution network, with values ​​ranging from 0 to 24 (integers). For distribution network numbering, For the dispatch period, DSO represents the distribution network side.

[0078] Distribution network For example, taking uncertain variables In the uncertain set The TS-RO model is constructed with the goal of minimizing operating costs when the internal environment changes towards the "worst" scenario.

[0079]

[0080] Wherein, the outermost min represents the first-stage problem, and the optimization variable is... The inner max-min problem is the second-stage problem, and the optimization variables are: and In the formula This is the coefficient vector matrix corresponding to each constraint condition when converted to a compact form. A constant column vector, This is the column vector of coefficients corresponding to the objective function. As shown in the following formula:

[0081]

[0082] The two-phase problem of TS-RO is decomposed into a main problem and subproblems. The main problem after decomposition is:

[0083] in, and These are the current iteration number and the maximum iteration number, respectively. For the first Solution to the subproblem after the next iteration For the first In the worst-case scenario after the next iteration The value of .

[0084] The decomposed subproblems are:

[0085]

[0086] In the formula: Indicates a given set hour The feasible region is expressed as follows:

[0087]

[0088] in, , , , These are the dual variables corresponding to each constraint in the second-stage min problem.

[0089] Strong duality theory can transform the inner-layer min problem into a max problem, and then merge it with the outer-layer max problem into a single overall max problem. The transformation process is as follows:

[0090] The inner min problem in the subproblem is:

[0091]

[0092] In the given The minimization problem of the lower inner layer can be transformed into a linear programming problem.

[0093] (1) Introduce dual variables:

[0094] Introduced dual variables Corresponding to equality constraints that only relate to variables in the second stage. Corresponding to inequality constraints that only relate to the variables in the second stage. Corresponding to inequality constraints that relate to variables in both stages 1 and 2. The corresponding constraints are that the values ​​of wind and solar power output and load power are given uncertain variables. The value of .

[0095] (2) By introducing dual variables, it can be rewritten in the form of a Lagrange function:

[0096]

[0097] (3) Simplify the Lagrange function:

[0098]

[0099] (4) Lagrange dual function:

[0100]

[0101] If you want If the coefficient matrix is ​​bounded, then the linear terms preceding it must satisfy:

[0102]

[0103] Right now:

[0104]

[0105] (5) Based on strong duality theory, the inner-level min problem is transformed into a max problem:

[0106]

[0107] (6) Combine the outer max problem with a single max problem to obtain the following subproblems:

[0108]

[0109] The form involves the product of binary and continuous variables, which is further linearized using the Big M method.

[0110]

[0111] in, The introduced continuous auxiliary variable has a value range of 10³ to 10⁻⁶. 6 It should be noted that the value of M should be appropriate, because if the value of M is too small, it may lead to infeasible solutions, and if the value of M is too large, it may cause the constraints to fail or the values ​​to be unstable due to computer precision issues.

[0112] After the above derivation and transformation, the TS-RO model is finally decoupled into a main problem and subproblems in mixed integer linear form, which are then solved using a column and constraint generation algorithm.

[0113] S2: Day-ahead global scheduling phase: Taking the worst-case operating scenario obtained in S1 as a deterministic input, and aiming to minimize the overall operating cost of the transmission and distribution network, a transmission and distribution coordinated optimization scheduling model based on the objective cascade method (ATC) is constructed. The transmission network and distribution network are decoupled through boundary interaction power. The transmission network equates the interaction power to virtual loads, and the distribution network equates the interaction power to virtual generators. After the transmission and distribution networks are solved independently, only the boundary power information is coordinated to obtain the day-ahead optimal scheduling strategy for transmission and distribution coordination.

[0114] The process of solving the ATC algorithm is as follows:

[0115] (1) The power exchange between transmission and distribution networks still needs to meet the consistency constraint:

[0116]

[0117] in, for The transmission network transmits to the first time period The interactive power of the distribution network for Time period The interactive power transmitted from the distribution network to the transmission network.

[0118] (2) The objective function of the transmission and distribution network during the daytime stage, in addition to the total operating cost during the daytime stage and the deviation penalty cost before the daytime stage, also needs to add the ATC penalty cost. This penalty cost aims to guide the interaction power between the transmission network and the distribution network to gradually approach each other during the calculation process, thereby achieving consistency.

[0119] The objective function for the power transmission network is modified as follows:

[0120]

[0121] The objective function for the distribution network k is modified as follows:

[0122]

[0123] in, The total operating cost of the power transmission network, For the first The total operating cost of a power distribution network For the number of distribution networks, and For the multiplier of the penalty term in the ATC algorithm, and The interaction power value is obtained from the adjacent region. The superscript "-" indicates that this item is a known quantity.

[0124] (3) By setting a penalty term, the interactive power of the transmission and distribution network gradually approaches the target value during the calculation process, thereby achieving consistency and satisfying the convergence requirement. The convergence condition is:

[0125]

[0126] in, and The convergence coefficient is . For the number of iterations, For the region No. The objective function value of the next iteration. For the region No. The objective function value of the next iteration.

[0127] If the convergence condition is not met, then update the penalty function factor:

[0128]

[0129] in, It is a constant. and The initial value is a relatively small constant. and The first The algorithm's penalty term multiplier in the next iteration, and These are the multipliers for the penalty term in the updated algorithm. For the first iteration The transmission network transmits to the first time period The interactive power of the distribution network For the first iteration Time period The interactive power transmitted from the distribution network to the transmission network.

[0130] S3: The process of embedding ATC into MPC:

[0131] The intraday transport and distribution coordinated optimization scheduling model embeds ATC into the rolling loop of MPC. In the day-ahead phase, ATC deals with a 24-hour global optimization problem, while in the intraday phase, due to the introduction of the rolling optimization mechanism, when time progresses to the [missing information], [missing information] During each scheduling period, the control platform obtains the current prediction domain. Real-time forecast data of wind and solar power output and load demand within the forecast domain are used, and the transmission and distribution networks within the forecast domain are decoupled through boundary interactive power. While optimizing their respective networks, interactive power information is transmitted, and transmission and distribution coordinated optimization is achieved through ATC. Only one operation is performed. The scheduling strategy for each time period. Next, the prediction domain advances one time period and continues optimization, while still executing the scheduling plan for the first time period. Through iteration, the prediction domain continues to advance, and the scheduling domain gradually moves along with the prediction domain. In the last 4 hours of the scheduling cycle, the prediction domain gradually decreases and continues until the scheduling plans for all time periods within the scheduling cycle are completed.

[0132] Example 2

[0133] Reference Figure 4-6 This embodiment will verify the implementation effect of the technical solution described in Embodiment 1:

[0134] Scenario example: Based on a 6-node transmission network and 2-node distribution network interconnection system (e.g.) Figure 1 The transmission network includes two conventional thermal power units, Unit 1 with a maximum output of 100MW and Unit 2 with a maximum output of 80MW, as well as hydropower units and wind power units. Each distribution network is equipped with an energy storage device with a rated capacity of 24MWh and a maximum charging / discharging power of 4.8MW; the number of dispatchable electric vehicles is 150, with an average charging / discharging power of 0.06MW per unit; the maximum interruptible load is 2MW, and the total transferable load capacity is 32MWh. Electricity pricing mechanism: peak-hour price is 930 yuan / MWh, normal-hour price is 580 yuan / MWh, and off-peak price is 250 yuan / MWh. ATC algorithm convergence coefficient. , Initial value of multipliers ; .

[0135] To fully verify the technical effects of this invention, three sets of experiments were set up. All experiments were conducted based on the same set of historical wind and solar load data to ensure the fairness of the comparison results.

[0136] Experiment 1: Compare the operating costs of "independent optimization of transmission and distribution" and "ATC collaborative optimization of this invention" to verify the global economy and robustness;

[0137] Experiment 2: Compare the operating costs of "day-ahead ATC collaborative optimization" and "day-ahead-intraday multi-timescale collaborative optimization" to verify the economic improvement effect of rolling correction;

[0138] Experiment 3: Test the performance of the method of this invention under different prediction error scenarios and verify its reliability in extreme scenarios.

[0139] Experiment 1:

[0140] This experiment is based on the worst-case operating scenario obtained from the TS-RO model. In this scenario, the wind and solar power output is at the lower limit of fluctuation and the load is at the upper limit of fluctuation, which is the most unfavorable operating condition for the system.

[0141] Independent transmission and distribution optimization mode: The transmission network and distribution network formulate scheduling plans completely independently, without any information exchange. The transmission network only arranges the output of thermal power units according to its own load demand, and the distribution network only relies on internal energy storage and demand response resources to balance power fluctuations;

[0142] This invention's ATC collaborative optimization mode decouples and coordinates the transmission and distribution networks through boundary power interaction, transmitting only boundary power information and not exchanging any internal sensitive data. The overall system operating cost is minimized through iterative solution using the ATC algorithm.

[0143] The operating cost calculation results for the two modes are shown in the table below:

[0144] Table 1 Comparison of Operating Costs of Independent and Collaborative Optimization of Transmission and Distribution Networks

[0145] Cost / yuan Distribution Network 1 Distribution Network 2 power transmission network Total cost Independent optimization 49963.39 50595.75 338442.84 439001.98 Collaborative optimization 49795.25 51275.31 314246.98 415317.54

[0146] The following conclusions can be drawn from the data in Table 1: a balance between system robustness and overall economy can be achieved.

[0147] After adopting the ATC distributed collaborative optimization of the present invention, the total operating cost of the system was reduced from RMB 439,001.98 to RMB 415,317.54, a reduction of 5.40%, and the overall economic efficiency was significantly improved.

[0148] The transmission network saw the largest decrease in operating costs, from 338,442.84 yuan to 314,246.98 yuan, a reduction of 7.15%. This is because the distribution network actively adjusts its internal flexible resources to provide auxiliary support to the transmission network, reducing the expensive start-up and shutdown costs of thermal power units and reserve capacity on the transmission network side.

[0149] In the independent optimization mode, the ability to flexibly exchange resources is fragmented. By adopting the ATC distributed collaborative optimization method of this invention, the total system cost is reduced by 5.40% under the premise of only interacting with boundary power (without privacy leakage) between the transmission and distribution networks. The distribution network significantly reduces the operational pressure on the transmission network through proactive support.

[0150] Experiment 2:

[0151] This experiment is based on real-time updated ultra-short-term wind and solar load forecast data, comparing the operating costs of two scheduling modes:

[0152] ATC Collaborative Optimization Based Only on the Most Severe Situation Scenario of the Present Day: The 24-hour scheduling plan is formulated based solely on the worst-case scenario of the present day, without any dynamic adjustments during the day, and is strictly executed according to the present day plan;

[0153] This invention proposes a multi-timescale collaborative optimization method for day-ahead and intraday scheduling: Based on the day-ahead scheduling plan, an MPC rolling optimization mechanism is introduced, with a 15-minute time scale and a 4-hour prediction time domain. The scheduling plan is dynamically corrected using real-time high-precision prediction data. At the same time, a penalty cost for energy storage deviating from the day-ahead plan is introduced into the intraday objective function.

[0154] The operating cost calculation results for the two modes are shown in the table below.

[0155] Table 2 Comparison of Day-ahead and Intraday Collaborative Optimization Operating Costs of Transmission and Distribution Networks

[0156] Cost / yuan Distribution Network 1 Distribution Network 2 power transmission network Total cost Current stage 49795.25 51275.31 314246.98 415317.54 Intraday phase 44999.79 43687.78 300837.73 389525.29

[0157] Results Analysis

[0158] The following conclusions can be drawn from the data in Table 2:

[0159] After introducing intraday MPC rolling correction, the total operating cost of the system decreased from RMB 415,317.54 to RMB 389,525.29, a decrease of 6.21% compared to the previous period, and the overall economic efficiency was improved;

[0160] The operating costs of the distribution network saw the most significant reduction, with distribution network 1 decreasing by 9.63% and distribution network 2 decreasing by 14.80%.

[0161] In the daytime phase, a large amount of reserve capacity was reserved to mitigate extreme risks, resulting in high costs. Entering the intraday phase, MPC rolling optimization was introduced, dynamically adjusting based on 15-minute high-precision forecast data, releasing conservative reserve capacity, and further reducing the total cost of intraday collaborative optimization by 6.21% compared to the daytime phase.

[0162] Experiment 3:

[0163] To verify the reliability of the method of this invention under extreme uncertainty scenarios, three test scenarios with different prediction errors were set up, covering the full range from normal operation to extremely harsh operation:

[0164] Real-time scenario: The wind and solar load forecasting error is controlled within a small range, corresponding to the normal operating conditions of the system;

[0165] Poor scenario: Increased wind and solar load forecasting errors, corresponding to harsh system operating conditions;

[0166] Extreme scenario: The wind and solar load forecasting error reaches its maximum, corresponding to extremely poor system operating conditions and extremely poor net load level of the distribution network.

[0167] The power supply and distribution coordinated scheduling algorithm of this invention was run in each scenario, and the system operating cost and whether load shedding occurred were recorded. The results are shown in the table below:

[0168] Table 3 Comparison of Operating Costs for Different Forecasting Scenarios During the Intraday Phase

[0169] Cost / yuan Real-time scene Poor scenario Extremely poor scene Distribution Network 1 44999.79 52391.54 63237.02 Distribution Network 2 43687.78 50670.27 59947.79 power transmission network 300837.73 317075.96 344845.76 Total cost 389525.29 420137.77 468030.57

[0170] The following conclusions can be drawn from the data in Table 3: Reliability in the face of extreme uncertainty scenarios.

[0171] Even in the most extreme operating scenarios, where the net load level of the distribution network is extremely poor, the algorithm of this invention can still converge quickly and achieve power balance. No load shedding occurred, effectively demonstrating the operability and high reliability of the model under various harsh operating conditions.

[0172] Summary of Results:

[0173] Operational robustness and cost-effectiveness: TS-RO identifies the worst-case scenarios to ensure safety boundaries, and ATC collaboration reduces the total cost by 5.40%;

[0174] Real-time and precise control capabilities: The intraday MPC rolling correction mechanism overcomes day-ahead forecast bias, further reducing operating costs by 6.21%;

[0175] Safety and reliability in extreme scenarios: Even in extreme scenarios, the algorithm can still converge quickly and achieve power balance.

[0176] Equipment lifespan protection: By introducing a deviation penalty function into the intraday model, frequent disordered charging and discharging caused by the energy storage device to smooth out fluctuations is avoided. Comparing the scheduling results with and without penalty costs: without penalty costs, the number of daily charging and discharging cycles of energy storage is 3.2 times, and the number of charging and discharging state switching times is 18 times / day; after introducing the dual penalty cost of this invention, the number of daily charging and discharging cycles of energy storage is reduced to 2.1 times, and the number of state switching times is reduced to 7 times / day, with a reduction of 34.4% in the number of cycles, which significantly extends the equipment lifespan.

[0177] In summary, the above description is merely a specific embodiment of the present invention, intended to help those skilled in the art understand or implement the present invention. Modifications to these embodiments will be readily apparent to those skilled in the art. The general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not limited to the embodiments shown herein, but should be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for coordinated economic dispatch of transmission and distribution considering a high proportion of distributed energy access, characterized in that, include: S1: Day-ahead uncertainty handling stage: Based on the day-ahead wind and solar load forecast data, a two-stage robust optimization TS-RO model containing the box uncertainty set of wind and solar load is constructed with a 1-hour time scale; With the objective of minimizing the system operating cost when the uncertain variables take the worst operating scenario within the uncertainty set, the TS-RO model is decomposed into a main problem and sub-problems of mixed integer linear form and solved to obtain the worst operating scenario of wind and solar load. S2: Day-ahead global scheduling phase: Taking the worst-case operating scenario obtained in S1 as a deterministic input, and aiming to minimize the overall operating cost of the transmission and distribution network, a transmission and distribution coordinated optimization scheduling model based on the objective cascade method (ATC) is constructed. The transmission network and distribution network are decoupled through boundary interaction power. The transmission network treats the interaction power as a virtual load, and the distribution network treats the interaction power as a virtual generator. After the transmission and distribution networks are solved independently, only the boundary power information is coordinated to obtain the day-ahead optimal scheduling strategy for transmission and distribution coordination. S3: Intraday Rolling Correction Phase: Introducing the rolling optimization concept of Model Predictive Control (MPC), with a 15-minute time scale and a 4-hour prediction time domain, the start-up and shutdown status of thermal power units and the basic operation plan of energy storage in the day-ahead optimal scheduling strategy obtained in S2 are used as deterministic constraints; using real-time updated ultra-short-term wind and solar load forecast data, an intraday phase multi-time scale rolling collaborative optimization model based on ATC is constructed, and a double penalty cost for energy storage deviating from the day-ahead plan is introduced into the intraday objective function; the optimal transmission and distribution coordination plan for the next 4 hours is solved through the ATC algorithm, and only the scheduling instructions for the first 15-minute period are executed. Then, the prediction time domain is shifted forward by one period, and the above process is repeated until the scheduling of all 96 periods of the day is completed.

2. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The aforementioned daytime TS-RO scheduling model specifically includes the construction of the box-shaped uncertainty set of wind and solar loads: To describe the uncertainty of wind and solar load, a box-shaped uncertainty set for the fluctuation range of wind and solar load is constructed, specifically represented as follows: Where U is the box-type uncertain set of wind and solar loads. For a set of uncertain variables, , , Distribution network During the period The values ​​of wind power output, photovoltaic power output, and load active power under uncertain scenarios, where B is a set of binary variables. , , These are 0-1 variables representing whether the uncertainties of wind power, solar power, and load have reached their boundaries. When the value is 1, the wind power and solar power output reaches their minimum values, and the load power reaches its maximum values ​​for the corresponding time period. This represents the maximum fluctuation deviation of wind power output in the distribution network. This represents the predicted wind power output of the distribution network. This represents the maximum fluctuation deviation of photovoltaic power output in the distribution network. This is the predicted value of photovoltaic output in the distribution network. This represents the maximum fluctuation deviation of the active power of the distribution network load. This represents the predicted value of the active power of the distribution network load. , , These are the uncertainty adjustment parameters for wind and solar loads introduced into the distribution network, with values ​​ranging from 0 to 24 (integers). For distribution network numbering, For the dispatch period, DSO represents the distribution network side.

3. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The solution process for the current daytime TS-RO scheduling model is as follows: By adjusting the uncertainty adjustment parameter The conservatism of the flexible adjustment and scheduling scheme; The strong duality theory is used to transform the min problem in the inner layer of the TS-RO model into a max problem. After merging with the max problem in the outer layer, the original model is decoupled into a main problem and subproblems in mixed integer linear form. The column and constraint generation algorithm (C&CG) is used to iteratively solve the main problem and subproblems until the convergence condition is met, thus obtaining the worst-case operating scenario.

4. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The ATC-based transmission and distribution coordination decoupling mechanism described in S2 is specifically: a day-ahead stage ATC-based transmission and distribution coordination optimization scheduling model, which achieves decoupling and coordination between the transmission network and the distribution network through boundary interactive power; The interactive power is equivalent to virtual generators on the distribution network side and virtual loads on the transmission network side, and consistency constraints are introduced: in, for The transmission network transmits to the first time period The interactive power of the distribution network for Time period The interactive power transmitted from the distribution network to the transmission network. By adding penalty terms based on the Lagrange multiplier method to the independent objective functions of the transmission and distribution networks, the reconstructed day-ahead ATC objective function is as follows: in, The total operating cost of the power transmission network, For the first The total operating cost of a power distribution network For the number of distribution networks, and For the multiplier of the penalty term in the ATC algorithm, and The interaction power value is obtained from the adjacent region. The superscript "-" indicates that the item is a known quantity.

5. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The total operating cost of the transmission and distribution network mentioned in S2 includes: the power generation cost, start-up and shutdown cost, reserve cost, renewable energy power generation cost, and transmission and distribution interaction cost of the thermal power units on the transmission network side; and the operating cost of energy storage devices, interruptible load compensation cost, transferable load adjustment cost, charging and discharging cost of electric vehicles using the V2G mode, renewable energy power generation cost, and transmission and distribution interaction cost of the distribution network side.

6. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The solution process for intraday rolling collaborative optimization described in S3 is as follows: In the current prediction time domain Within the current forecast domain, the system acquires 15-minute ultra-short-term wind and solar load forecast data. Based on the current forecast data, iteratively solves the intraday ATC model until the interactive power of the transmission and distribution network meets the convergence condition, and outputs the multi-period coordinated scheduling results within the current forecast domain. Extract the optimal scheduling instruction for the first 15-minute period and issue it for execution. Then, set t=t+1, predict the time domain and slide it forward by one period. Repeat the above optimization process until the scheduling of all 96 periods of the day is completed.

7. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The expression for the double penalty cost of energy storage deviating from the day-ahead plan, as described in S3, is as follows: Where f is the penalty cost. This is the intraday period. For the intraday rolling optimization of the prediction time domain, and The penalty cost for energy storage devices during the daytime phase of the distribution network, and The penalty coefficient is... and For the intraday phase of the distribution network Energy storage charging and discharging status during a given period of time. and This represents the day-ahead charging and discharging status of energy storage in the distribution network. and For the intraday phase of the distribution network Energy storage charging and discharging power during a given period and This represents the day-ahead energy storage charging and discharging power of the distribution network.

8. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: The day-ahead ATC model of S2 and the intraday ATC model of S3 adopt the same convergence criterion and multiplier update rule. The convergence condition is as follows: in, and The convergence coefficient is . For the number of iterations, For the region No. The objective function value of the next iteration. For the region No. The objective function value of the next iteration. If the convergence condition is not met, then update the penalty function factor: in, It is a constant. and The initial value is a relatively small constant. and The first The algorithm's penalty term multiplier in the next iteration, and These are the multipliers for the penalty term in the updated algorithm. For the first iteration The transmission network transmits to the first time period The interactive power of the distribution network For the first iteration Time period The interactive power transmitted from the distribution network to the transmission network.

9. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: In S2, the distribution network forms an adjustable capacity boundary through the joint dispatch of various flexible resources within its internal system, specifically: The power distribution network coordinates the scheduling of energy storage devices, interruptible loads, transferable loads, and electric vehicles using the V2G mode to calculate the upper limit of adjustable capacity and the lower limit of adjustable capacity. By constraining the power transmission and distribution interaction within the physically feasible range through the adjustable capacity boundary, the power interaction command that exceeds the limit is avoided.

10. The transmission and distribution coordinated economic dispatch method considering a high proportion of distributed energy access as described in claim 1, characterized in that: When converting the TS-RO model to a mixed-integer linear form, the product terms of binary and continuous variables in the model are linearized using the Big M method. The Big M coefficients for the auxiliary variables range from 10³ to 10⁻⁶. 6 .

Images