Substation capacity planning method and device, electronic equipment and medium

Abstract

The application relates to the technical field of power system planning, and discloses a substation capacity planning method and device, an electronic device and a medium; the method comprises the following steps: acquiring a data set in a planning area, and generating standard calculation format data; calculating the net load of each node based on the standard calculation format data, and screening target nodes in combination with a new energy reverse sending influence index system; for the target nodes, taking the node voltage amplitude square and phase angle as state variables, performing first-order Taylor expansion on the original alternating current flow equation, linearly replacing the active and reactive power network loss term by using the plane cutting method, and constructing a linearized alternating current flow constraint; a two-stage robust optimization model is established; the two-stage robust optimization model is iteratively solved to obtain a main transformer capacity selection and number configuration scheme of each target node; and a target planning scheme is output. The application solves the problems that it is difficult to cope with voltage out-of-limit caused by new energy reverse sending, reactive capacity is blocked, and energy loss is serious under the traditional voltage reduction power supply planning mode.

Description

Technical Field

[0001] This application relates to the field of power system planning technology, and more specifically, to a substation capacity planning method, device, electronic equipment, and medium. Background Technology

[0002] With the high proportion of distributed renewable energy integration in the new power system, the distribution network has transformed from a unidirectional power receiving network into an active network with bidirectional power exchange capabilities. During periods of peak renewable energy generation, power is routinely fed back to the 500kV main grid.

[0003] However, traditional 500kV main transformer planning often focuses on capacity configuration around peak load, without fully considering voltage rise and reactive power capacity bottlenecks caused by reverse power transmission. This makes it easy for planning schemes to trigger excessive reactive power transmission when facing large-scale renewable energy generation, leading to large-scale power curtailment and hindering the local consumption of renewable energy. Summary of the Invention

[0004] In view of the above situation, this application provides a substation capacity planning method, device, electronic equipment and medium, which aims to solve the above problems or at least partially solve the above problems.

[0005] In a first aspect, embodiments of this application provide a substation capacity planning method, the method comprising: Acquire load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and perform data cleaning, verification and format conversion to generate standard calculation format data; The net load of each node is calculated based on the standard calculation format data, and target nodes are selected in conjunction with the new energy back-transmission impact index system. For the target node, the original AC power flow equation is expanded in first order using the square of the node voltage amplitude and the phase angle as state variables. The active and reactive power loss terms are linearized by using the cutting plane method to construct linearized AC power flow constraints. Using the linearized AC power flow constraints, the PQ power factor coupling constraints of the new energy grid connection point, the reactive power back-feeding capacity constraints of the main transformer, the reverse setting constraints of the relay protection, and the short-term overload capacity constraints as constraints, a two-stage robust optimization model is established with the goal of minimizing the comprehensive cost over the entire life cycle. The dual column constraint generation algorithm is used to iteratively solve the two-stage robust optimization model to obtain the main transformer capacity selection and number configuration scheme for each target node; Output the target planning scheme, which includes the selection and configuration scheme of the main transformer capacity of the target node, the economic indicators of the whole life cycle and the system operation verification results.

[0006] Secondly, embodiments of this application also provide a substation capacity planning device, the device comprising: The acquisition module is used to acquire load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and to perform data cleaning, verification and format conversion to generate standard calculation format data. The filtering module is used to calculate the net load of each node based on the standard calculation format data, and to filter target nodes in conjunction with the new energy back-transmission impact index system. The linearized power flow construction module is used to perform a first-order Taylor expansion of the original AC power flow equation for the target node, with the square of the node voltage magnitude and the phase angle as state variables, and to use the cutting plane method to linearize the active and reactive power loss terms to construct linearized AC power flow constraints. The robust planning model construction module is used to establish a two-stage robust optimization model with the constraints of linearized AC power flow, PQ power factor coupling constraint of new energy grid connection point, reactive power back-feed capacity constraint of main transformer, reverse setting constraint of relay protection and short-term overload capacity constraint as constraints, and with the goal of minimizing the comprehensive cost of the whole life cycle. The solution module is used to iteratively solve the two-stage robust optimization model using a dual column constraint generation algorithm to obtain the main transformer capacity selection and number configuration scheme for each target node. The output module outputs the target planning scheme, which includes the selection and configuration scheme of the main transformer capacity of the target node, the economic indicators of the whole life cycle, and the system operation verification results.

[0007] Thirdly, embodiments of this application also provide an electronic device, including: a processor; and a memory arranged to store computer-executable instructions, which, when executed, cause the processor to perform the steps described in the first aspect.

[0008] Fourthly, embodiments of this application also provide a computer-readable storage medium that stores one or more programs, which, when executed by an electronic device including multiple applications, cause the electronic device to perform the steps described in the first aspect.

[0009] The at least one technical solution adopted in this application embodiment can achieve the following beneficial effects: By calculating the net load and using the new energy backfeed impact index system, the node most significantly affected by power backfeed can be accurately identified, avoiding investment waste caused by indiscriminate expansion of the entire network; By performing a first-order Taylor expansion on the original AC power flow equation and using plane-cutting linearization, a completely linear AC power flow constraint is constructed, eliminating the non-convexity of the traditional power flow equation, enabling large-scale mixed integer programming to be solved within an acceptable time, supporting the planning calculation of actual provincial power grids; The two-stage robust optimization model adopted does not need to rely on the accurate probability distribution of new energy output to ensure the safety constraints under all feasible scenarios, overcoming the defect of deterministic programming causing scheme instability due to probability distribution deviation; The model contains The PQ coupling constraints, reactive power back-feed capacity constraints, and relay protection reverse setting and overload constraints directly solve practical engineering problems such as reactive power bottlenecks, voltage overruns, and relay protection malfunctions caused by renewable energy back-feeding. The dual column constraint generation algorithm transforms the three-layer problem into a two-layer problem through dual transformation of subproblems, significantly reducing the number of variables and the scale of constraints. At the same time, it achieves an adaptive balance between investment decisions and operational risks through cutting plane iteration. The final output planning scheme includes the configuration of main transformer models and number of units at each node, full life cycle economic indicators, and operational verification results, directly serving equipment bidding and engineering construction decisions. It effectively improves the grid's carrying capacity for renewable energy back-feeding at multiple voltage levels, the renewable energy absorption rate, and the power supply safety level under extreme faults, and has significant engineering application value. Attached Figure Description

[0010] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating the substation capacity planning method provided in an embodiment of this application is shown. Figure 2 A schematic diagram of the power transmission network topology provided in an embodiment of this application is shown; Figure 3 A structural diagram of the substation capacity planning device provided in an embodiment of this application is shown; Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0012] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such use can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the term "comprising" and its variations should be interpreted as open-ended terms meaning "including but not limited to."

[0013] Figure 1 This paper illustrates a flowchart of the substation capacity planning method provided in an embodiment of this application. Figure 1 It can be seen that this application includes at least steps S101-S106: Step S101: Obtain load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and perform data cleaning, verification and format conversion to generate standard calculation format data.

[0014] The load data includes: historical and projected annual load curves, load simultaneity rate, and power factor for each 500kV node. Power source data includes the installed capacity, upper and lower limits of output, generation costs, and output curves for external DC scenarios for conventional units (coal-fired, gas-fired, pumped-storage, and nuclear power); information on inter-provincial interconnection lines; and the installed capacity, typical daily output curves, and power factor adjustment range for wind power and photovoltaic power at voltage levels of 220kV, 110kV, and below. Transmission network data includes: the 500kV grid topology, line numbers, starting and ending nodes, length, resistance, reactance, susceptance, and rated transmission capacity; and the capacity, years of operation, and operating status of existing 500kV transformers. The transformer technical parameters include: parameters for the mainstream domestic ODFS / ODFPS series 500kV three-phase autotransformers, including unit cost, full life-cycle design life, relay protection limitations, and reverse load limitations for three models: 750MVA (single-phase 250MVA), 1000MVA (single-phase 334MVA), and 1200MVA (single-phase 400MVA). Cost parameters include: load shedding penalty cost, discount rate, and annual fixed operation and maintenance fee rate for the transformer.

[0015] In some embodiments, data cleaning, verification, and format conversion include: removing outliers and missing values ​​from the original data and supplementing missing time-series data; verifying the matching of load and power capacity, the consistency of line parameters, and the industry compliance of cost parameters; and converting the cleaned and verified data into a standard matrix format.

[0016] Specifically, the raw time-series data may contain extreme and unreasonable values ​​due to measurement failures or manual input errors, such as negative load values ​​or sudden changes exceeding the normal fluctuation range by several times. These outliers must be directly removed. For data gaps caused by communication interruptions, etc., the average of preceding and following times or interpolation using similar daily data should be used to supplement the data to ensure its integrity. Further checks should be made to ensure that the load forecast values ​​for each node are consistent with the regional economic development plan, and that the installed capacity of new energy sources does not exceed a reasonable multiple of the region's maximum load. If they do not match, it indicates data errors or that planning targets need adjustment. The parameters such as transmission line length, resistance, reactance, and susceptance should be checked to ensure they are consistent with the State Grid's general design standards to avoid deviations in power flow calculation results due to parameter input errors. The unit cost of transformers and penalty costs should be checked to ensure they are within a reasonable range in the current market, ensuring the objectivity of the economic evaluation. Finally, the cleaned and verified raw tabular data should be converted into a standard matrix format that can be directly called by subsequent optimization solvers (such as Gurobi).

[0017] The standard matrix format includes a node-branch correlation matrix, a generator node correlation matrix, a typical scenario output matrix, and a fault scenario state matrix. The node-branch correlation matrix describes which lines connect to which nodes, used to construct network topology constraints; the generator node correlation matrix describes which node each conventional unit is connected to, used to construct output constraints; the typical scenario output matrix organizes the typical daily output curves and load curves of wind and solar power into a matrix form according to scenarios, used for the second stage of operation simulation; the fault scenario state matrix: pre-enumerates N-1 and N-2 fault conditions, encodes their states into a 0-1 matrix, used for extreme scenario verification.

[0018] In some embodiments, the node-branch correlation matrix is ​​transformed as follows: Based on the first and last node numbers of each transmission line in the line table, a two-dimensional array containing the first and last node numbers of all lines is generated, and the node susceptance matrix is ​​calculated using the line reactance parameters. The diagonal elements of the matrix are the sum of the admittances of all lines connected to the node, and the off-diagonal elements are the negative values ​​of the admittances of the lines connecting the two nodes, forming a sparse coefficient matrix that directly participates in the linearized power flow constraint.

[0019] The conversion method of the generator node association matrix is ​​as follows: Based on the 500kV node number to which each conventional unit belongs, a logical matrix consisting of 0 and 1 is generated. The number of rows in the matrix is ​​the total number of nodes, and the number of columns is the total number of units. If a unit is connected to a node, the corresponding position in the matrix is ​​set to 1, otherwise it is set to 0. This is used to accumulate the output variables of each unit to the power balance equation of the corresponding node according to the node to which it belongs.

[0020] The conversion method of the typical scenario output matrix is ​​as follows: For the load curve, wind power output curve and photovoltaic power output curve of 8760 hours throughout the year, the clustering algorithm or the selection of typical days according to the season is used to compress the scenario and generate a time series coefficient matrix containing a finite number of typical scenarios. Each scenario contains the per-unit load value or actual power value at 24 times, which is used as the known external parameter input in the second stage of operation simulation.

[0021] The transformation method of the fault scenario state matrix is ​​as follows: Based on the pre-generated set of anticipated faults, a 0-1 matrix containing the operating state codes of each line under each of the N-1 or N-2 fault scenarios is generated. This matrix participates in the dynamic correction of network topology coefficients during the solution process and is used for safety constraint verification under extreme fault scenarios. Each matrix is ​​stored in the form of a sparse or dense numerical array, which can be directly called by the optimization solver during model construction.

[0022] Step S102: Calculate the net load of each node based on the standard calculation format data, and select target nodes in conjunction with the new energy back-transmission impact index system.

[0023] Net load represents the power that the node actually needs to absorb from the 500kV main grid or the power it needs to feed back to the main grid at this moment.

[0024] In some embodiments, the node load time series data and the lower-level multi-voltage level renewable energy output time series data of typical scenarios in the planning level year are extracted from the standard calculation format data. The net load power of each 500kV node is calculated time by time, and the minimum net load of each node in the whole scenario is extracted. Nodes with minimum net load less than zero are selected to form a candidate node set. Various indicators of the renewable energy back-transmission impact index system are calculated for the candidate node set. After comprehensive evaluation, nodes that are significantly affected by back-transmission are selected as target nodes.

[0025] Specifically, for each node, at each moment in a typical scenario, the net load at that moment is obtained by subtracting the total output of the downstream renewable energy sources at that moment from the load value at that moment. This process iterates through all scenarios and all moments to find the minimum net load value for each node across all moments. This minimum value represents the node's performance during the most severe backfeeding event; the more negative the value, the more severe the backfeeding. All nodes with negative minimum net loads are marked as a candidate node set. The candidate nodes are further refined by ranking the calculated indicators from highest to lowest risk, selecting nodes with the highest comprehensive scores, and simultaneously eliminating nodes with minor backfeeding impacts based on engineering experience thresholds. This ultimately forms a candidate set of target nodes for priority expansion or new construction of main transformers.

[0026] The indicator system for the impact of renewable energy back-transmission includes the proportion of renewable energy back-transmission power at nodes, node voltage sensitivity, average load rate of associated lines, and renewable energy penetration rate in the downstream distribution network.

[0027] Specifically, the reverse power ratio is calculated by dividing the absolute value of the minimum net load of the node by its existing main transformer rated capacity. The resulting percentage measures the impact of reverse power on the existing main transformer capacity; a higher ratio indicates greater pressure on the existing main transformer to bear the reverse power. Node voltage sensitivity is calculated by performing a small-disturbance simulation on the node, slightly changing the total output of renewable energy sources under the node, and observing the change in the 500kV bus voltage amplitude. The ratio of the two is the voltage sensitivity; a higher sensitivity indicates that the node's voltage is more severely affected by renewable energy fluctuations and is more prone to voltage overruns due to reverse power transmission. The average load factor of associated lines is calculated by statistically analyzing the load factors of all 500kV transmission lines directly connected to the node during periods of high renewable energy generation, and taking the arithmetic mean. A higher value indicates a more significant crowding-out effect of reverse power on surrounding transmission channels and a greater risk of transmission obstruction. The penetration rate of new energy in the lower-level distribution network is calculated by dividing the total installed capacity of wind power and photovoltaic power at voltage levels of 220kV and below within the power supply area of ​​the node by the maximum load value of the area. The resulting ratio is used to characterize the scale and level of new energy development in the area. A higher penetration rate indicates a greater potential for backflow and a higher possibility of exacerbating the problem.

[0028] Furthermore, the calculation results of the four indicators are sorted from highest to lowest risk level. For each candidate node, its ranking position in each of the four indicators is taken, and a weighted comprehensive score is calculated according to preset weights. The weights can be flexibly set according to the actual focus of the power grid in different regions. For example, if the transmission channel in a region is a major bottleneck, the weight of the average load rate of the associated lines can be appropriately increased; if the voltage support in a region is weak, the weight of voltage sensitivity can be increased. Finally, the candidate nodes are sorted in descending order of comprehensive score.

[0029] Furthermore, two screening rules are established. The first is a quantity rule, which selects the top-ranked nodes based on their comprehensive scores according to the annual investment budget, such as the top 30% or top 20. The second is a threshold rule, which mandates the inclusion of nodes with extremely high individual performance indicators, even if their comprehensive scores are not high. For example, nodes with a reverse power ratio exceeding 90% or voltage sensitivity exceeding a set safety threshold must be unconditionally included in the target node set. Combining these two rules forms the final candidate set of target nodes.

[0030] In addition, the selected target node candidate set can be handed over to engineering planners for manual review to check whether there are any nodes that were mistakenly selected due to data anomalies, or whether there are any nodes that, although they have not reached the screening threshold, need to be included due to their strong regional importance. After review and adjustment, the target node set for subsequent two-stage robust optimization modeling is finally locked.

[0031] Step S103: For the target node, with the square of the node voltage magnitude and the phase angle as state variables, perform a first-order Taylor expansion on the original AC power flow equation, and use the cutting plane method to linearize the active and reactive power loss terms to construct a linearized AC power flow constraint.

[0032] In some embodiments, the square of the voltage amplitude and the voltage phase angle of each node are used as state variables. The node injection active power equation and reactive power equation of the original power flow equation are respectively expanded by first-order Taylor expansion. The trigonometric function terms and voltage cross terms in the expansion are linearly approximated to obtain a preliminary linearized expression for the node injection power. A network loss auxiliary variable is introduced, and the nonlinear quadratic product terms corresponding to active network loss and reactive network loss are respectively substituted to transform the nonlinear quadratic terms in the original equation into the linear expression of the auxiliary variable. Multiple sets of linear cutting plane inequalities are constructed using the cutting plane linearization method to externally approximate the non-convex network loss nonlinear terms, and all nonlinear constraints related to network loss are transformed into linear inequality constraints.

[0033] Specifically, the original power flow equations are as follows:

[0034]

[0035] With voltage phase angle Square of voltage amplitude As state variables, a Taylor expansion of the original power flow equations is performed, and the following approximation is made:

[0036] We can obtain a preliminary linearized expression for the node injection power:

[0037]

[0038] In the above formula Approximately ,Will Approximately And for nonlinear terms Perform the following transformations:

[0039] Combining the above substitutions, we can obtain:

[0040]

[0041] in:

[0042]

[0043] Using the same line of thought, we can further obtain:

[0044]

[0045] The AC power flow node injection power equation, after the above assumptions and variable substitutions, still contains nonlinear quadratic terms in its first-order Taylor expansion. By linearizing through a cutting plane, the non-convex network loss terms are transformed into a set of inequality constraints. Let: ,

[0046] And by:

[0047] have:

[0048] The nonlinear term of reactive power loss is handled in the same way and will not be elaborated further.

[0049] The following linearization of the cutting plane yields the following network loss linearity inequality constraint:

[0050]

[0051] In the above formula, Represents a node The injected active power, Represents a node With nodes Mutual conductance between them Represents a node voltage amplitude, Represents a node voltage amplitude, Represents a node With nodes voltage phase angle difference, Represents a node With nodes Mutual susceptance between them Represents a node The injected reactive power, where n represents the total number of nodes in the system. This represents the value of the tangent point corresponding to the m-th cutting plane.

[0052] In summary, by integrating node injection power balance constraints, inter-node phase angle difference constraints, node voltage magnitude constraints, and network loss linear inequality constraints, a complete set of linearized AC power flow constraints is formed, while preserving the linear coupling relationship between active power, reactive power, and node voltage magnitude.

[0053] This model eliminates trigonometric functions and quadratic coupling terms through linear approximation, transforming the power flow equations into a system of purely linear algebraic equations. This significantly improves the efficiency of solving mixed-integer programming while preserving the linear coupling relationship between active power P, reactive power Q, and the square of the node voltage magnitude U.

[0054] Step S104: Using linearized AC power flow constraints, PQ power factor coupling constraints at new energy grid connection points, reactive power feedback capacity constraints of main transformers, reverse setting constraints of relay protection, and short-term overload capacity constraints as constraints, a two-stage robust optimization model is established with the goal of minimizing the comprehensive cost over the entire life cycle.

[0055] In some embodiments, the two-stage robust optimization model adopts a two-stage decision-making framework: the first stage is the investment decision-making stage, where the decision variables are the model and number of main transformers at each target node; the second stage is the operation verification stage, where, based on the investment decision, operation simulation optimization is performed for typical operating scenarios and N-2 extreme failure scenarios, with the goal of minimizing the overall cost over the entire life cycle.

[0056] The objective equation for optimization is as follows:

[0057]

[0058]

[0059]

[0060]

[0061] in, For transformer capacity investment cost, For transformer operation and maintenance costs, This represents the operating costs under typical scenarios, including the cost of conventional power generation and the cost of energy curtailment losses from renewable energy sources. This represents the minimum load shedding and energy curtailment cost under extreme scenarios and the most severe failures. A collection of typical scenarios, A collection of extreme scenarios. Let be the set of times t in the scene. Let s be the number of days in a year when the s-th typical scenario occurs. The number of days in a year when the s-th extreme scenario occurs is calculated by multiplying the probability of the s-th typical or extreme scenario by the number of days in a year. This gives the number of days in a year when the corresponding scenario occurs. This represents the uncertain set of the most severe N-2 faults in a transformer under extreme scenarios. Investment cost for selecting the capacity of a 500kV transformer. For the set of power transmission network nodes, This is a collection of capacities for different types of transformers. Select a transformer of type k for node i. Choose to build n units for node i. For single unit capacity (750 / 1000 / 1200MVA), The cost is per unit. 1.8% is the annual fixed operation and maintenance fee rate for the transformer. For operating costs, For operation and maintenance costs. Cost per unit capacity of the generating unit The output of unit g at time t in scenario s. It is a collection of energy storage nodes. For the unit operating cost of energy storage, Energy storage charging power, Energy storage discharge power. For pumped storage unit node set, Contribute to the anticipated scenarios of new energy, The active power is fed back to the current. This is consumed by the load of the lower-level power grid. This is a set of nodes in a 500kV transmission network. This refers to the cost of load shedding and energy curtailment under the most severe fault conditions in extreme scenarios. The cost of load shedding per unit capacity For the cost of energy curtailment, This refers to the load shearing magnitude.

[0062] Core constraints: 500kV node power balance constraints: , .

[0063] In the formula, A collection of conventional generating units on a power transmission network node. For the set of power transmission network nodes, Let the active power output of unit j on node i of the transmission network be at time t under typical scenario s. The active power transmitted back to node i. Let the reactive power output of unit j on transmission network node i at time t under typical scenario s be given. The reactive power transmitted back to node i. This refers to the reactive power compensation capacity configured on node i.

[0064] Linearized power flow constraints: Constraints constructed in step S103.

[0065] New energy reverse transmission PQ coupling constraints: Among them, the power factor at its rated operating point is selected. Expand the calculation.

[0066] Main transformer reactive power feedback capacity constraint: , .in, The reactive power compensation ratio is a key indicator for measuring the matching relationship between the capacitive reactive power compensation capacity and the core equipment capacity of a substation. Its inductive reactive power compensation capacity is usually configured at 20% to 25% of the main transformer capacity.

[0067] Integrated feedback limit of relay protection and load rate: .

[0068] Overall load capacity ratio constraint: .

[0069] In the formula, The 0-1 variables of node i for constructing type k transformer, Configure the number of transformers on node i. For the newly constructed main transformer capacity of node i; The original total capacity of the 500kV main transformer in this area The highest predicted load for node i in the planning year. Power factor (usually taken as 0.95). , Upper and lower limits of capacity ratio.

[0070] Step S105: The dual column constraint generation algorithm is used to iteratively solve the two-stage robust optimization model to obtain the main transformer capacity selection and number configuration scheme for each target node.

[0071] In some embodiments, auxiliary variables and the Big M method are introduced to linearize the bilinear and nonlinear terms in the model; the two-stage robust optimization model is decomposed into a main problem and subproblems, with the main problem solving for investment decision variables and the subproblems solving for the operating cost under the worst-case scenario; dual transformation is performed on the lower-level subproblems to transform the minimization problem into a maximization problem; a dual column constraint generation algorithm is used for iterative solution, with the cutting plane generated by the subproblems added to the main problem in each iteration, until the difference between the objective functions of the main problem and the subproblems is less than the convergence threshold of 0.5%; the calculation is completed by calling the YALMIP modeling toolkit and the mixed integer programming solver.

[0072] To address the bilinear and nonlinear terms in the model, a dual column constraint generation algorithm is used for solution. The specific execution steps are as follows: Model linearization: For the bilinear terms of transformer type and quantity, an auxiliary variable is introduced and constraints are added, where M is a sufficiently large positive number (taken as 10000): , , , .

[0073] All nonlinear network loss terms are processed using the cutting plane linearization method.

[0074] Model decomposition: The original two-stage robust optimization model is decomposed into a main problem (MP) and subproblems (SP): Main problem: Solve for investment decision variables, with the goal of minimizing the sum of investment costs and operating costs corresponding to the currently known cutting plane.

[0075] Subproblem: Given an investment decision, solve for the maximum operating cost in the worst-case scenario and generate a new cutting plane.

[0076] Dual transformation: The problem of minimizing the running cost of the lower-level subproblems is transformed into a dual transformation to obtain the maximization problem, which transforms the original three-level Min-Max-Min problem into a two-level Max-Min problem, reducing the number of variables by about 30%.

[0077] Iterative solution: Initialization: The number of iterations k=0, and the initial set of cutting planes for the main problem is empty.

[0078] Step 1: Solve the main problem to obtain the investment decision and the objective function value LB (lower bound).

[0079] Step 2: Substitute the planning results from the first stage into the subproblem to solve for the worst-case operating cost UB (upper bound).

[0080] Step 3: If the difference between the upper and lower bounds is less than the convergence threshold of 0.5%, the iteration terminates and the optimal solution is output; otherwise, the cutting plane generated by the subproblem is added to the main problem, k=k+1, and the process returns to step 1.

[0081] Solver call: The YALMIP R2023a toolkit was used for modeling, and the Gurobi 9.5 solver was called with MIPGap=0.005.

[0082] Step S106: Output the target planning scheme.

[0083] The target planning scheme includes the selection of main transformer capacity and configuration of the number of units for the target nodes, the economic indicators of the whole life cycle and the system operation verification results.

[0084] For example, the main transformer planning scheme includes: the model and number of main transformers at each node, and the capacity of new / expanded construction; economic indicators include: annual value of main transformer investment, annual operation and maintenance costs, annual power generation costs, annual energy curtailment costs, annual load shedding costs, and total life cycle costs; operational indicators include: comprehensive energy curtailment rate, comprehensive load shedding rate, average load rate of main transformers, and maximum energy curtailment and load shedding under extreme fault conditions; and visualization charts include: typical daily main transformer power transmission curves, spatiotemporal distribution maps of load rates on associated lines, and distribution maps of regional renewable energy absorption capacity.

[0085] from Figure 1 As can be seen from the method described, this application accurately identifies the nodes most significantly impacted by power backflow through net load calculation and a new energy backflow impact index system, avoiding investment waste caused by indiscriminate expansion of the entire network; by performing a first-order Taylor expansion of the original AC power flow equations and using plane-cutting linearization, a completely linear AC power flow constraint is constructed, eliminating the non-convexity of the traditional power flow equations, enabling large-scale mixed integer programming to be solved within an acceptable time, supporting the planning calculation of actual provincial power grids; the two-stage robust optimization model adopted does not rely on the accurate probability distribution of new energy output to ensure safety constraints in all feasible scenarios, overcoming the defect of deterministic programming causing scheme instability due to probability distribution deviation; the PQ coupling constraint incorporated in the model The system addresses practical engineering problems caused by reactive power back-feeding capacity constraints and relay protection reverse setting and overload constraints, directly resolving issues such as reactive power bottlenecks, voltage overruns, and relay protection malfunctions. The dual-column constraint generation algorithm transforms a three-layer problem into a two-layer problem through dual transformation of subproblems, significantly reducing the number of variables and the scale of constraints. Simultaneously, it achieves an adaptive balance between investment decisions and operational risks through cutting plane iteration. The final output planning scheme includes the main transformer model and number configuration for each node, full life-cycle economic indicators, and operational verification results, directly serving equipment bidding and engineering construction decisions. This effectively improves the power grid's capacity to handle multi-voltage-level renewable energy back-feeding, the renewable energy absorption rate, and the power supply safety level under extreme faults, demonstrating significant engineering application value.

[0086] Furthermore, in some embodiments of this application, if the current output target planning scheme shows in the preset verification indicators that the average load rate of the main transformer of at least one target node exceeds the threshold or the comprehensive curtailment rate does not meet the requirements of the regulations, then the net load screening threshold value of the target node is lowered, and the target node screening step is re-executed; the re-screened target node is substituted into the two-stage robust optimization model for re-solution until the operation verification indicators of all target nodes meet the engineering qualification standards, and the final planning scheme that can guide equipment bidding and engineering construction is output.

[0087] In this embodiment, traditional planning methods typically end after outputting a solution, with the selection threshold relying entirely on manual experience and determined once. If the threshold is set too strictly, leading to missed key nodes, or too loosely, resulting in an excessively large optimization model that is difficult to solve, automatic correction is not possible afterward. However, this invention introduces a post-output operational verification and feedback mechanism. When key indicators such as the average load rate of the main transformer or the comprehensive energy curtailment rate do not meet the engineering qualification standards, the net load selection threshold for the corresponding target node is automatically narrowed, and the selection and optimization are re-executed. This achieves a dynamic closed loop from "one-time static selection" to "verification-feedback-re-selection-re-optimization," effectively solving the problem of... Improper initial screening thresholds can lead to omissions of key nodes or substandard planning schemes. Meanwhile, this feedback mechanism eliminates the need for engineers to rely on extensive trial calculations to find the optimal screening threshold in the early stages of planning. Instead, the system adaptively approximates qualified solutions during iterations, significantly reducing the workload of repeated manual parameter adjustments. Furthermore, each iteration uses actual operational verification results as the basis for correction, ensuring that the final output planning scheme meets load rate and energy curtailment rate requirements at all target nodes. This effectively guarantees that the planning scheme can be directly used to guide equipment bidding and engineering construction, avoiding later design changes and project delays due to substandard schemes, and significantly improving the engineering practicality and robustness of the planning method.

[0088] The following is combined Figure 2 The schematic diagram of the power transmission network topology shown illustrates the scheme of this application.

[0089] Figure 2It includes 164 500kV nodes and 342 500kV transmission lines; the total installed capacity of conventional power sources is 144,803MW, including coal-fired power, gas turbine power, hydropower, pumped storage and nuclear power; the total installed capacity of wind power is 45,694MW and the total installed capacity of photovoltaic power is 124,863MW, of which 85% are connected to voltage levels of 220kV and below; the maximum load of the entire grid is 172,042.03MW. The cost of energy curtailment penalty is 0.05 million yuan / MWh, and the cost of load shedding penalty is 2.5 million yuan / MWh; the unit investment for a 750MVA transformer is 150,000 yuan / MVA, for a 1000MVA transformer it is 140,000 yuan / MVA, and for a 1200MVA transformer it is 130,000 yuan / MVA; the transformer's lifespan is 25 years, the discount rate is 8%, and the annual fixed operation and maintenance cost rate is 1.8%; the grid connection point power factor is 0.95, the upper limit of reactive power feedback ratio of the main transformer is 20%, the upper limit of short-term load rate is 85%, the N-1 overload factor is 1.5, and the capacity-to-load ratio range is 1.5~2.1. The calculation results of the minimum net load index of key nodes are shown in Table 1; the model adopts the dual column constraint generation algorithm and is solved using the YALMIP toolkit and the Gurobi solver, with a solution accuracy set to 0.5%. The results of the model's economic and technical indicators are shown in Table 2, and the transformer capacity selection results are shown in Table 3.

[0090] Table 1 Calculation results of minimum net load index for key nodes

[0091] Table 2 Results of the model's solution for economic and technical indicators

[0092] Table 3 Transformer Capacity Selection Results

[0093] The solution results from the example show that there are 164 500kV nodes. Ignoring the planning of 1050kV nodes, the planning results indicate that 5 nodes will have new transformers built, and 32 nodes will be expanded, with a total new / expanded transformer capacity of 74650MWA. Based on the aforementioned transformer transmission capacity and line load conditions, and considering the relevant transformer reactive power compensation configuration, it can be seen that, considering transformer capacity selection, transformer transmission capacity is no longer the limiting factor for energy curtailment and load shedding. More importantly, it's the limitations imposed by lines and main transformer transmission reaching their limits, leading to the inability to deliver renewable energy in a timely manner, resulting in energy curtailment. Considering other constraints on transformers, such as relay protection and reverse load rate limitations, PQ coupling constraints, and overall capacity-to-load ratio, will improve the renewable energy absorption rate and effectively reduce load shedding.

[0094] Compared with existing technologies, the planning method proposed in this application can effectively balance the investment cost of 500kV main transformers and the capacity for renewable energy absorption, significantly improve the grid's capacity to support the power backflow of renewable energy at multiple voltage levels, and at the same time ensure power supply safety under extreme fault conditions and improve the capacity for renewable energy absorption, which has important engineering application value.

[0095] In some embodiments of this application, a substation capacity planning device is provided, which corresponds one-to-one with the substation capacity planning method described in the above embodiments. For example... Figure 3 As shown, the substation capacity planning device includes an acquisition module 101, a filtering module 102, a linearized power flow construction module 103, a robust planning model construction module 104, a solution module 105, and an output module 106. The acquisition module 101 is used to acquire load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and to perform data cleaning, verification and format conversion to generate standard calculation format data. The screening module 102 is used to calculate the net load of each node based on the standard calculation format data, and to screen target nodes in conjunction with the new energy back-transmission impact index system. The linearized power flow construction module 103 is used to perform a first-order Taylor expansion of the original AC power flow equation for the target node, with the square of the node voltage magnitude and the phase angle as state variables, and to use the cutting plane method to linearize the active and reactive power loss terms to construct linearized AC power flow constraints. The robust planning model construction module 104 is used to establish a two-stage robust optimization model with the constraints of linearized AC power flow, PQ power factor coupling constraint of new energy grid connection point, reactive power back-feeding capacity constraint of main transformer, reverse setting constraint of relay protection and short-term overload capacity constraint as constraints, and with the goal of minimizing the comprehensive cost of the whole life cycle. The solution module 105 is used to iteratively solve the two-stage robust optimization model using a dual column constraint generation algorithm to obtain the main transformer capacity selection and number configuration scheme for each target node. The output module 106 is used to output the target planning scheme, which includes the selection and configuration scheme of the main transformer capacity of the target node, the economic indicators of the whole life cycle and the system operation verification results.

[0096] In some embodiments of this application, in the above-mentioned apparatus, the acquisition module is specifically used to remove outliers and missing values ​​from the original data, supplement missing time-series data; verify the matching of load and power capacity, the consistency of line parameters, and the industry compliance of cost parameters; and convert the cleaned and verified data into a standard matrix format, wherein the standard matrix format includes a node-branch association matrix, a generator node association matrix, a typical scenario output matrix, and a fault scenario state matrix.

[0097] In some embodiments of this application, in the above-mentioned apparatus, the screening module 102 is specifically used to extract the node load time series data and the lower-level multi-voltage level new energy output time series data of typical scenarios in the planning level year from the standard calculation format data, calculate the net load power of each 500kV node at each time step, and extract the minimum net load of each node in the whole scenario. Nodes with a minimum net load less than zero are selected to form a candidate node set; For the candidate node set, various indicators of the new energy back-transmission impact index system are calculated. After comprehensive evaluation, nodes that are significantly affected by back-transmission are selected as target nodes. The new energy back-transmission impact index system includes the proportion of new energy back-transmission power of the node, the node voltage sensitivity, the average load rate of the associated lines, and the new energy penetration rate of the lower-level distribution network.

[0098] In some embodiments of this application, in the above-described apparatus, the linearized power flow construction module is specifically used to perform a first-order Taylor expansion of the node injection active power equation and reactive power equation of the original power flow equation, using the square of the voltage amplitude and the voltage phase angle of each node as state variables, and to perform linear approximation on the trigonometric function terms and voltage cross terms in the expansion to obtain a preliminary linearized node injection power expression. By introducing auxiliary variables for network loss, the nonlinear quadratic product terms corresponding to active power loss and reactive power loss are substituted, and the nonlinear quadratic terms in the original equation are transformed into linear expressions of the auxiliary variables. Multiple sets of linear cutting plane inequalities are constructed using the cutting plane linearization method to externally approximate the nonlinear terms of non-convex network loss, thereby transforming all nonlinear constraints related to network loss into linear inequality constraints. By integrating node injection power balance constraints, inter-node phase angle difference constraints, node voltage magnitude constraints, and network loss linear inequality constraints, a complete set of linearized AC power flow constraints is formed.

[0099] In some embodiments of this application, the two-stage robust optimization model in the above-described apparatus employs a two-stage decision framework: The first stage is the investment decision-making stage, and the decision variables are the model and number of main transformers configured at each target node; The second phase is the operational verification phase. Based on the investment decision, operational simulation optimization is carried out for typical operational scenarios and N-2 extreme failure scenarios, with the goal of minimizing the overall cost over the entire life cycle.

[0100] In some embodiments of this application, in the above-described apparatus, the solution module is specifically used to linearize the bilinear and nonlinear terms in the model by introducing auxiliary variables and the Big M method. The two-stage robust optimization model is decomposed into a main problem and sub-problems. The main problem solves for investment decision variables, while the sub-problems solve for the operating costs under the worst-case scenario. By performing a dual transformation on the lower-level subproblems, the minimization problem is transformed into a maximization problem; The dual column constraint generation algorithm is used for iterative solution. In each iteration, the cutting plane generated by the subproblem is added to the main problem until the difference between the objective functions of the main problem and the subproblem is less than the convergence threshold of 0.5%. The calculation is completed by calling the YALMIP modeling toolkit and the mixed integer programming solver.

[0101] In some embodiments of this application, in the above-described apparatus, the optimization module is used to reduce the net load screening threshold value of the target node and re-execute the target node screening step if the currently output target planning scheme shows in the preset verification indicators that at least one target node's average load rate of the main transformer exceeds the threshold or the comprehensive curtailment rate does not meet the requirements of the procedure. The re-selected target nodes are substituted into the two-stage robust optimization model for re-solution until the operational verification indicators of all target nodes meet the engineering qualification standards, and the final planning scheme that can guide equipment bidding and engineering construction is output.

[0102] It should be noted that any of the above-mentioned substation capacity planning devices can implement the aforementioned substation capacity planning method one by one, which will not be elaborated here.

[0103] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Figure 4 As shown, at the hardware level, this electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or it may include non-volatile memory, such as at least one disk drive. Of course, this electronic device may also include other hardware required for other business operations.

[0104] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0105] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0106] The processor reads the corresponding computer program from non-volatile memory into main memory and then runs it, forming a substation capacity planning device at the logical level. The processor executes the program stored in memory and specifically performs the aforementioned methods.

[0107] The processor may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0108] This electronic device can execute the substation capacity planning method provided in several embodiments of this application, and is implemented as a substation capacity planning device. Figure 3 The functions of the embodiments shown are not described in detail here.

[0109] This application also proposes a computer-readable storage medium that stores one or more programs, the programs including instructions that, when executed by an electronic device including multiple applications, enable the electronic device to perform the substation capacity planning method provided in several embodiments of this application.

[0110] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0111] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0112] 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.

[0113] 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.

[0114] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0115] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0116] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0117] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0118] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0119] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A substation capacity planning method, characterized in that, The method includes: Acquire load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and perform data cleaning, verification and format conversion to generate standard calculation format data; The net load of each node is calculated based on the standard calculation format data, and target nodes are selected in conjunction with the new energy back-transmission impact index system. For the target node, the original AC power flow equation is expanded in first order using the square of the node voltage amplitude and the phase angle as state variables. The active and reactive power loss terms are linearized by using the cutting plane method to construct linearized AC power flow constraints. Using the linearized AC power flow constraints, the PQ power factor coupling constraints of the new energy grid connection point, the reactive power back-feeding capacity constraints of the main transformer, the reverse setting constraints of the relay protection, and the short-term overload capacity constraints as constraints, a two-stage robust optimization model is established with the goal of minimizing the comprehensive cost over the entire life cycle. The dual column constraint generation algorithm is used to iteratively solve the two-stage robust optimization model to obtain the main transformer capacity selection and number configuration scheme for each target node; Output the target planning scheme, which includes the selection and configuration scheme of the main transformer capacity of the target node, the economic indicators of the whole life cycle and the system operation verification results.

2. The method according to claim 1, characterized in that, The process of data cleaning, verification, and format conversion to generate standard computational format data includes: Remove outliers and missing values ​​from the original data and fill in the missing time series data; Verify the matching of load and power capacity, the consistency of line parameters, and the industry compliance of cost parameters; The cleaned and verified data is converted into a standard matrix format, which includes a node-branch correlation matrix, a generator node correlation matrix, a typical scenario output matrix, and a fault scenario state matrix.

3. The method according to claim 1, characterized in that, The calculation of the net load of each node based on the standard calculation format data, and the selection of target nodes in conjunction with the new energy backflow impact index system, includes: Extract the node load time series data and the lower-level multi-voltage level new energy output time series data of typical scenarios in the planning level year from the standard calculation format data, calculate the net load power of each 500kV node at each time step, and extract the minimum net load of each node in the whole scenario. Nodes with a minimum net load less than zero are selected to form a candidate node set; For the candidate node set, various indicators of the new energy back-transmission impact index system are calculated. After comprehensive evaluation, nodes that are significantly affected by back-transmission are selected as target nodes. The new energy back-transmission impact index system includes the proportion of new energy back-transmission power of the node, the node voltage sensitivity, the average load rate of the associated lines, and the new energy penetration rate of the lower-level distribution network.

4. The method according to claim 1, characterized in that, The method involves performing a first-order Taylor expansion of the original AC power flow equations using the squares of the node voltage magnitudes and phase angles as state variables, and then employing the cutting plane method to linearize the active and reactive power loss terms to construct linearized AC power flow constraints, including: Using the square of the voltage amplitude and the voltage phase angle of each node as state variables, the node injection active power equation and reactive power equation of the original power flow equation are expanded by first-order Taylor, and the trigonometric function terms and voltage cross terms in the expansion are linearly approximated to obtain the preliminary linearized expression of node injection power. By introducing auxiliary variables for network loss, the nonlinear quadratic product terms corresponding to active power loss and reactive power loss are substituted, and the nonlinear quadratic terms in the original equation are transformed into linear expressions of the auxiliary variables. Multiple sets of linear cutting plane inequalities are constructed using the cutting plane linearization method to externally approximate the nonlinear terms of non-convex network loss, thereby transforming all nonlinear constraints related to network loss into linear inequality constraints. By integrating node injection power balance constraints, inter-node phase angle difference constraints, node voltage magnitude constraints, and network loss linear inequality constraints, a complete set of linearized AC power flow constraints is formed.

5. The method according to claim 1, characterized in that, The two-stage robust optimization model adopts a two-stage decision framework: The first stage is the investment decision-making stage, and the decision variables are the model and number of main transformers configured at each target node; The second phase is the operational verification phase. Based on the investment decision, operational simulation optimization is carried out for typical operational scenarios and N-2 extreme failure scenarios, with the goal of minimizing the overall cost over the entire life cycle.

6. The method according to claim 1, characterized in that, The iterative solution using the dual column constraint generation algorithm includes: By introducing auxiliary variables and the Big M method, the bilinear and nonlinear terms in the model are linearized. The two-stage robust optimization model is decomposed into a main problem and sub-problems. The main problem solves for investment decision variables, while the sub-problems solve for the operating costs under the worst-case scenario. By performing a dual transformation on the lower-level subproblems, the minimization problem is transformed into a maximization problem; The dual column constraint generation algorithm is used for iterative solution. In each iteration, the cutting plane generated by the subproblem is added to the main problem until the difference between the objective functions of the main problem and the subproblem is less than the convergence threshold of 0.5%. The calculation is completed by calling the YALMIP modeling toolkit and the mixed integer programming solver.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: If the current output target planning scheme shows that at least one target node's average load rate of the main transformer exceeds the threshold or the comprehensive energy curtailment rate does not meet the requirements of the regulations, then the net load screening threshold value of the target node will be lowered and the target node screening step will be re-executed. The re-selected target nodes are substituted into the two-stage robust optimization model for re-solution until the operational verification indicators of all target nodes meet the engineering qualification standards, and the final planning scheme that can guide equipment bidding and engineering construction is output.

8. A substation capacity planning device, characterized in that, The device includes: The acquisition module is used to acquire load data, multi-voltage level power supply data, transmission network topology and parameters, transformer technical parameters and cost parameters within the planning area, and to perform data cleaning, verification and format conversion to generate standard calculation format data. The filtering module is used to calculate the net load of each node based on the standard calculation format data, and to filter target nodes in conjunction with the new energy back-transmission impact index system. The linearized power flow construction module is used to perform a first-order Taylor expansion of the original AC power flow equation for the target node, with the square of the node voltage magnitude and the phase angle as state variables, and to use the cutting plane method to linearize the active and reactive power loss terms to construct linearized AC power flow constraints. The robust planning model construction module is used to establish a two-stage robust optimization model with the constraints of linearized AC power flow, PQ power factor coupling constraint of new energy grid connection point, reactive power back-feed capacity constraint of main transformer, reverse setting constraint of relay protection and short-term overload capacity constraint as constraints, and with the goal of minimizing the comprehensive cost of the whole life cycle. The solution module is used to iteratively solve the two-stage robust optimization model using a dual column constraint generation algorithm to obtain the main transformer capacity selection and number configuration scheme for each target node. The output module is used to output the target planning scheme, which includes the selection and configuration scheme of the main transformer capacity of the target node, the economic indicators of the whole life cycle and the system operation verification results.

9. An electronic device, comprising: processor; as well as A memory is configured to store computer-executable instructions, which, when executed, cause the processor to perform the steps of the substation capacity planning method as described in any one of claims 1-7.

10. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of applications, cause the electronic device to perform the steps of the substation capacity planning method as described in any one of claims 1-7.

Images