Collaborative planning method and system for county power distribution network and local micro-grid of power system

Abstract

The invention discloses a cooperative planning method for a county power distribution network and a local micro-grid of a power system. The method comprises the following steps: acquiring data information of the county power distribution network and the local micro-grid of a target power system; constructing a cooperative planning operation scene of the county power distribution network and the local micro-grid of the target power system; constructing a collaborative planning model of the county power distribution network and the local micro-grid of the power system; decomposing the constructed model to construct a layered and distributed solution model; and the model is solved based on an ADMM algorithm, and collaborative planning of the county power distribution network and the local micro-grid of the target power system is completed. The invention also discloses a system for realizing the cooperative planning method for the county power distribution network and the local micro-grid of the power system. Collaborative planning of the county power distribution network and the local micro-grid of the power system is realized, the reliability is higher, and the accuracy is better.

Description

Technical Field

[0001] This invention belongs to the field of electrical automation, specifically relating to a collaborative planning method and system for county-level power distribution networks and local microgrids in a power system. Background Technology

[0002] With economic and technological development and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and daily life, bringing endless convenience. Therefore, ensuring a stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] Currently, with the increasing severity of environmental problems, more and more new energy power generation systems are being integrated into the power grid and generating electricity. Therefore, county-level power grids are becoming the access points for distributed photovoltaic, wind power, and other new energy power generation systems. Thus, appropriate planning for county-level power distribution networks is of great significance for ensuring the safe and stable operation of the power system and county-level power grids.

[0004] Currently, power system planning schemes for county-level power grids still employ traditional, standardized conventional distribution network planning schemes. However, county-level power grids differ from conventional distribution networks in their unique characteristics, load features, and grid infrastructure, and the scale of renewable energy sources also differs. Therefore, conventional distribution network planning schemes are not suitable for county-level power grid planning. Furthermore, although some researchers have proposed planning schemes for county-level power grids, these schemes do not consider the characteristics of renewable energy generation systems or the characteristics of corresponding local microgrids, resulting in relatively low reliability and accuracy. Summary of the Invention

[0005] One of the objectives of this invention is to provide a highly reliable and accurate collaborative planning method for county-level power distribution networks and local microgrids in power systems.

[0006] The second objective of this invention is to provide a system for implementing the collaborative planning method of the county-level power distribution network and local microgrids in the power system.

[0007] The collaborative planning method for county-level power distribution networks and local microgrids provided by this invention includes the following steps:

[0008] S1. Obtain data information on the county-level distribution network and local microgrids of the target power system;

[0009] S2. Based on the data obtained in step S1, construct a collaborative planning and operation scenario for the county-level power distribution network and local microgrids of the target power system;

[0010] S3. Based on the operational scenario constructed in step S2, construct a collaborative planning model for the county-level power distribution network and local microgrids in the power system;

[0011] S4. Decompose the model constructed in step S3 to construct a hierarchical distributed solution model;

[0012] S5. Based on the ADMM algorithm, solve the model obtained in step S4 to complete the collaborative planning of the target power system county distribution network and local microgrid.

[0013] Step S2, which involves constructing a collaborative planning and operation scenario for the target power system's county-level distribution network and local microgrids based on the data obtained in step S1, specifically includes the following steps:

[0014] Planning cycle The decision is divided into K consecutive decision-making stages; the starting time of each stage k is... ;

[0015] Based on historical data and a pre-trained large model, several random paths are generated for each uncertainty factor, and each path represents a possible future evolution trajectory.

[0016] The generated paths are used to construct an initial scene tree T; each node n in the initial scene tree T represents a possible state at the corresponding time point; a path from the root node to a leaf node of the initial scene tree is called a scene s, with a corresponding probability of... ;

[0017] The initial scene tree is reduced using the following steps:

[0018] Calculate any two scenarios s and s using the following formula. Distance between :

[0019] In the formula Let s be the vector of uncertainty variables in scenario s at stage k, including load. Photovoltaic cost Wind power costs Light intensity and wind speed ; It is the 2-norm of the vector, i.e., the Euclidean norm; Let be the overall volatility index for scenario s, and , Let be the standard deviation of each uncertainty variable in the k-th stage; Let be the average correlation coefficient among different uncertainty variables in scenario s; The set fluctuation weight; The relevant weights are set;

[0020] Calculate all possible pairs of scenes s and Distance between And construct the corresponding distance matrix. for ;

[0021] Hierarchical clustering algorithm is used for scene clustering;

[0022] For each class The scene with the smallest average distance from other scenes was selected as the representative scene. ;

[0023] The probability representing the scenario Set as ;

[0024] After the above steps, the initial n scenarios are reduced to M representative scenarios and their corresponding probabilities. These representative scenarios constitute the reduced scenario set, which is used for subsequent collaborative planning and optimization. This reduction process significantly reduces the scenario size and improves computational efficiency while preserving the main statistical features of the original scenario tree.

[0025] Complete the construction of a collaborative planning and operation scenario for the county-level distribution network and local microgrids of the target power system.

[0026] Step S3, which involves constructing a collaborative planning model for the county-level power distribution network and local microgrids based on the operational scenario established in step S2, specifically includes the following steps:

[0027] Construct the objective function:

[0028] The following formula is used as the objective function:

[0029] In the formula The set investment cost weighting coefficient; For investment costs, and , For a distributed set of devices, Let i be the capital recovery factor of equipment i and , The annual discount rate is 10%. For the economic service life of device i, Let i be the unit capacity investment cost of equipment. For the additional construction capacity of device i in phase s of scenario s, Sunk costs for reserving capacity per unit of equipment i. For scenario s, stage k represents the capacity reserved for device i. For the collection of distribution network lines, Let be the capital recovery coefficient for line j. Let j be the investment cost per unit length of the line. The new or upgraded length of line j in stage s of scenario s; The set weighting coefficient for operation and maintenance costs; For operation and maintenance costs, and , The total number of time periods within stage k. The duration of a single time step. Let t be the electricity price purchased from the upper-level power grid at time t. Let k be the power purchased by k from the upstream power grid at time t in scenario s. For gas turbine assembly, The unit fuel cost of a gas turbine. For the active power output of device i at time t in phase s of scenario s, The unit maintenance cost of device i. This refers to the collection of all distributed power devices in the system. The system risk cost weighting coefficient is set; For systemic risk costs, and , For the system node set, Let m be the load value of node m. Let be the load reduction amount at time t for node k and node m in scenario s. This is the penalty factor for exceeding the voltage limit per unit. For scenario s, node k and node n at time t, the voltage exceeds the limit and , For reference voltage, This represents the maximum permissible voltage deviation. The set weighting coefficient value for green energy consumption benefits; For the benefit of green energy consumption, and , The environmental benefits generated by green electricity per unit. For the actual photovoltaic output of k in phase s at time t, The actual wind power output of k in phase s at time t. The cost of curtailing wind and solar power per unit Let k be the amount of photovoltaic power curtailment at time t in phase s. Let k be the wind power curtailment power at time t in stage s. The duration of each planning phase;

[0030] Construct constraints:

[0031] Active power balance constraints:

[0032] In the formula Let n be the set of all power generation devices connected to node n; Let t be the active power flowing from the upper-level power grid into node n at time t in stage s. Let k be the active power flowing from node n to the upper-level power grid at time t in stage s. Let k be the load active power of node n at time t in phase s of scenario s; Let k be the load reduction amount at node n in phase s at time t. Let n be the set of all charging devices connected to node n. Let be the charging power of energy storage device j at time t in stage k of scenario s;

[0033] Reactive power balance constraints:

[0034] In the formula This is the set of all reactive power compensation devices connected to node n; The reactive power output of device i at time t in phase s of scenario s; Let K be the reactive power flowing from the upper-level power grid into node N at time t in stage s. Let K be the reactive power flowing from node N to the upper-level power grid at time t in stage s. Let K be the reactive power of the load at node n at time t in phase s of scenario s. The set of lines connected to node n; For scenario s, stage k at time t, the path The reactive power consumed;

[0035] Renewable energy output constraints:

[0036] In the formula The rated installed capacity of equipment i is defined as wind turbines and photovoltaics. Let k be the normalized output coefficient of device i at time t in stage s of scenario s;

[0037] Gas turbine operating constraints:

[0038] In the formula To contribute the minimum technical power to the gas turbine; To contribute the most technological expertise to gas turbines; Let k be the start / stop state of the gas turbine in scenario s at time t. If the gas turbine is in the start state in scenario s at time t, then... If, in scenario s, the gas turbine is in a non-starting state at time t, then... ; This represents the maximum ramp power of the gas turbine.

[0039] Energy storage system operating constraints:

[0040] In the formula Let k be the state of charge of the energy stored in scenario s at time t. For energy storage charging efficiency; The discharge efficiency of energy storage; The charging power of energy stored by k in stage s at time t; Let k be the discharge power of the energy stored in k at time t during stage s of scenario s; This refers to the rated capacity of the energy storage. This represents the minimum state of charge allowed for energy storage. This represents the maximum state of charge allowed for energy storage. This represents the maximum charging and discharging power of the energy storage.

[0041] Voltage quality constraints:

[0042] In the formula Let n be the voltage amplitude at node n at time k in stage s. The set lower voltage limit, and , The specified lower limit of the basic voltage. To ensure that the set reliability requirements reach the lower limit of the voltage range, The set risk weight coefficient; The set upper limit of voltage, and , The upper limit of the basic voltage is specified. The upper limit of the voltage range required to meet the set reliability requirements;

[0043] Power supply reliability constraints:

[0044] In the formula Total load power; For the maximum allowable expected power shortage, and , Allowed by the established economic orientation Maximum expected power shortage and , Allowed for the set high reliability requirements Maximum expected power shortage and ; The load reduction of node n at time k in phase s; This refers to the risk weighting coefficient.

[0045] Isolated operation capability constraints:

[0046] In the formula It is the set of all dispatchable distributed power sources within a microgrid; For scenario s at time k The total power of critical loads that must be guaranteed by the microgrid; Based on risk weight coefficient The defined continuous operation time requirement for microgrid islanding, and , The minimum self-balancing time is set. The set maximum self-balancing time.

[0047] Step S4, which involves decomposing the model constructed in step S3 to build a hierarchical distributed solution model, specifically includes the following steps:

[0048] The model constructed in step S3 is decomposed into an upper-layer model and a lower-layer model;

[0049] The upper-level model is for distribution networks, and the lower-level model is for microgrids;

[0050] Upper-level model:

[0051] The following formula is used as the objective function of the upper-level model:

[0052] In the formula The investment cost of the distribution network in stage k scenario s; The operating cost of the distribution network in scenario s at stage k; The active power price exchanged between the distribution network and the microgrid at time t in scenario k. This refers to the active power exchanged between the distribution network and the microgrid at time t in scenario k. The distribution network and microgrid exchange reactive power prices at time t in scenario k. This refers to the reactive power exchanged between the distribution network and the microgrid at time t in scenario k. The number of years included in each planning phase;

[0053] Upper-level model constraints:

[0054] Current constraints:

[0055] In the formula Let n be the set of lines ending at node n. This is the set of lines starting at node n; Let be the active power of line l at time t in stage k of scenario s; Let be the current value of line l at time t in stage k of scenario s; Let be the resistance of line l; Let n be the load power of node n at time t in stage k of scenario s; For the set of microgrids connected to node n;

[0056] Voltage coupling constraint:

[0057] In the formula Let be the voltage at connection point m of the microgrid at time t in stage k of scenario s; The voltage at connection point m on the distribution network side at time t in scenario s stage k;

[0058] Line capacity constraints:

[0059] In the formula This represents the maximum transmission capacity of line l. The set line capacity adaptive coefficient;

[0060] Lower-level model:

[0061] The following formula is used as the objective function of the lower-level model:

[0062] In the formula Let m be the investment cost of microgrid m in stage k of scenario s; Let m be the operating cost of microgrid m at time t in stage k of scenario s;

[0063] Lower-level model constraints:

[0064] Power balance constraints within a microgrid:

[0065] In the formula A collection of distributed generation equipment within a microgrid; The active power output of device i at time t in phase k of scenario s; Let m be the energy storage charging power of microgrid m at time t in scenario s phase k. Let be the load reduction amount of microgrid m at time t in phase k of scenario s;

[0066] Equipment operating constraints:

[0067] In the formula Minimum technical contribution to device i; The rated capacity of device i;

[0068] Reliability constraints:

[0069] In the formula The maximum allowable expected power shortage ratio for microgrid m;

[0070] Voltage consistency constraints:

[0071] In the formula Let be the voltage at connection point m on the m-side of the microgrid at time t during scenario s phase k.

[0072] Define the data exchange rules between the upper-level model and the lower-level model:

[0073] The data exchanged between the upper-layer model and the lower-layer model includes power information, voltage information, and price information; among which, power information includes... and Voltage information includes Price information includes and .

[0074] Step S5, which involves solving the model obtained in step S4 using the ADMM algorithm, includes the following steps:

[0075] The ADMM algorithm used is represented as follows: st

[0076] In the formula The objective function is the sub-problem of optimizing the upper-level distribution network. The sum of the objective functions of all the lower-level microgrid optimization subproblems; and The coefficient matrix describes the coupling relationships within the system; This represents the decision variable vector of the upper-level distribution network; This is the set of decision variables for all lower-level microgrids; This is the right-hand side term of the linear consistency constraint; This represents a linear consistency constraint;

[0077] The iterative process includes solving the x-subproblem, solving the z-subproblem, and updating the dual variable:

[0078] The solution to the x-subproblem is expressed as:

[0079] In the formula These are the decision variables for the distribution network side obtained in the (k+1)th iteration; Let x represent the augmented Lagrangian function with respect to the variable x; These are the microgrid-side decision variables obtained in the k-th iteration; Let be the dual variable of the k-th iteration;

[0080] The solution to the z-subproblem is expressed as:

[0081] In the formula These are the microgrid-side decision variables obtained in the (k+1)th iteration;

[0082] The update of the dual variable is represented as:

[0083] In the formula For the dual variable updated in the (k+1)th iteration; This is the penalty parameter used in the k-th iteration, and , As the first adjustment factor set, The second adjustment factor is set. For the original residual and , For dual residuals and , This is the 2-norm of a vector, also known as the Euclidean norm. The set balance threshold;

[0084] The augmented Lagrangian function is defined as follows:

[0085] In the formula For the k-th iteration, the augmented Lagrangian function is used to coordinate the decomposed distribution network and microgrid subproblems. Represents a weighted matrix The square of the weighted Euclidean norm is used to balance the penalty for violations of constraints on different physical dimensions (such as power and voltage). These are the values ​​of the distribution network-side decision variables obtained in the k-th iteration; The square of the weighted Euclidean norm with respect to the positive definite matrix P is used to measure the change of the distribution network variable x. These are the values ​​of the microgrid-side decision variables obtained in the k-th iteration; The weighted Euclidean norm square with respect to the positive definite matrix Q is used to measure the change of the microgrid variable z; Let be the dynamic regularization parameter at the k-th iteration, and , Initial regularization parameters, It is the attenuation factor, and ;

[0086] Map the model obtained in step S4 to the ADMM algorithm framework: map the objective function of the upper-level model to... The objective function of the lower-level model is corresponding to Linear consistency constraints include power balance constraints, voltage consistency constraints, and data exchange rules between the upper and lower level models; These are the decision variables for the higher-level model. For the decision variables of the lower-level model, c includes the power exchange boundary and voltage reference value;

[0087] At the beginning of each planning cycle, the ADMM algorithm initializes x, z, and y using the optimal solution from the previous stage; the state variables between planning cycles are passed through Markov chains.

[0088] For the lower-level model, an asynchronous parallel solution strategy is adopted: each microgrid independently solves its own z-subproblem, without considering the iteration completion status of other microgrids; the upper-level model, after collecting a set number of results from the lower-level model, solves its own x-subproblem to reduce synchronization waiting time; the asynchronous update rule is expressed as: for microgrid m, the corresponding z-subproblem is solved using the latest and most usable y and x; the upper-level model, after receiving .... After obtaining the results of the lower-level model, solve its own x-subproblems;

[0089] The standardized residuals are set as follows:

[0090] In the formula To standardize the original residual, it is used to measure the degree of violation of coupling constraints by the current iterative solution. The smaller the value, the better the physical constraints such as the overall power balance and voltage consistency of the system are satisfied. To standardize the dual residuals, which measure the drastic change in the dual variables, a smaller value indicates a more stable consensus in the distributed optimization process.

[0091] Set the maximum number of iterations to ;

[0092] Set the iteration termination condition:

[0093] Condition 1: ;in The first tolerance threshold is set;

[0094] Condition 2: ;in This is the set second tolerance threshold;

[0095] Condition 3: The number of iterations has been reached. ;

[0096] The ADMM algorithm used stops iterating when any one of condition 1, condition 2, or condition 3 is met.

[0097] The solution described in step S5 specifically includes the following steps:

[0098] (1) Initialization: Model parameter initialization;

[0099] (2) The lower-level model is solved in parallel: each microgrid independently optimizes its own scheme;

[0100] (3) Solving the upper-level model: Each distribution network optimizes its own scheme;

[0101] (4) Coordination variable update;

[0102] (5) Adjust algorithm parameters and optimize model performance;

[0103] (6) Convergence judgment; output the final planning scheme.

[0104] This invention also provides a system for implementing the collaborative planning method of the county-level distribution network and local microgrids of the power system, including a data acquisition module, a scenario construction module, a model construction module, a model decomposition module, and a collaborative planning module; the data acquisition module, scenario construction module, model construction module, model decomposition module, and collaborative planning module are connected in series; the data acquisition module is used to acquire data information of the county-level distribution network and local microgrids of the target power system and upload the data information to the scenario construction module; the scenario construction module is used to construct a collaborative planning operation scenario of the county-level distribution network and local microgrids of the target power system based on the received data information and the acquired data information, and upload the data information to the model construction module; the model construction module is used to construct a collaborative planning model of the county-level distribution network and local microgrids of the power system based on the received data information and the constructed operation scenario, and upload the data information to the model decomposition module; the model decomposition module is used to decompose the constructed model based on the received data information to construct a hierarchical distributed solution model, and upload the data information to the collaborative planning module; the collaborative planning module is used to solve the obtained model based on the ADMM algorithm based on the received data information to complete the collaborative planning of the county-level distribution network and local microgrids of the target power system.

[0105] The present invention provides a collaborative planning method and system for county-level power distribution networks and local microgrids in a power system. By acquiring and analyzing data information of the county-level power distribution network and local microgrids of the target power system to construct an operation scenario, and constructing and solving the corresponding collaborative planning model, it not only realizes the collaborative planning of county-level power distribution networks and local microgrids in a power system, but also has higher reliability and better accuracy. Attached Figure Description

[0106] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0107] Figure 2 This is a schematic diagram of the functional modules of the system of the present invention. Detailed Implementation

[0108] like Figure 1 The diagram shown is a flowchart of the method of the present invention: The collaborative planning method for county-level power distribution networks and local microgrids provided by the present invention includes the following steps:

[0109] S1. Obtain data information on the county-level distribution network and local microgrids of the target power system;

[0110] S2. Based on the data obtained in step S1, construct a collaborative planning and operation scenario for the target power system's county-level distribution network and local microgrids; specifically including the following steps:

[0111] Planning cycle The decision is divided into K consecutive decision-making stages; the starting time of each stage k is... ;

[0112] Based on historical data and pre-trained large models, several random paths are generated for each uncertain factor (such as load, photovoltaic energy storage investment cost, etc.), and each path represents a possible future evolution trajectory;

[0113] The generated paths are used to construct an initial scene tree T; each node n in the initial scene tree T represents a possible state at the corresponding time point; a path from the root node to a leaf node of the initial scene tree is called a scene s, with a corresponding probability of... ;

[0114] This invention employs an improved Kantorovich distance clustering algorithm. This algorithm not only considers the Euclidean distance of uncertain variables along each scene path, but also fully considers the fluctuation characteristics of the time series and the correlation between different variables. In specific implementation, the initial scene tree is reduced using the following steps:

[0115] Calculate any two scenarios s and s using the following formula. Distance between :

[0116] In the formula Let s be the vector of uncertainty variables in scenario s at stage k, including load. Photovoltaic cost Wind power costs Light intensity and wind speed ; It is the 2-norm of the vector, i.e., the Euclidean norm; Let be the overall volatility index for scenario s, and , Let be the standard deviation of each uncertainty variable in the k-th stage; Let be the average correlation coefficient among different uncertainty variables in scenario s; The set fluctuation weight; The relevant weights are set;

[0117] Calculate all possible pairs of scenes s and Distance between And construct the corresponding distance matrix. for ;

[0118] Hierarchical clustering algorithm is used for scene clustering;

[0119] For each class The scene with the smallest average distance from other scenes was selected as the representative scene. ;

[0120] The probability representing the scenario Set as ;

[0121] After the above steps, the initial n scenarios are reduced to M representative scenarios and their corresponding probabilities. These representative scenarios constitute the reduced scenario set, which is used for subsequent collaborative planning and optimization. This reduction process significantly reduces the scenario size and improves computational efficiency while preserving the main statistical features of the original scenario tree.

[0122] Complete the construction of a collaborative planning and operation scenario for the county-level distribution network and local microgrids of the target power system;

[0123] By using the above scenario reduction methods, we can retain those "key scenarios" that may not be extreme in numerical terms but have very different fluctuation patterns, thus more comprehensively covering various forms of uncertainty and improving the robustness of stochastic programming. At the same time, by considering the correlation between variables, we can ensure that the reduced scenario tree can truly reflect the intrinsic relationship between uncertainty factors.

[0124] S3. Based on the operational scenario constructed in step S2, construct a collaborative planning model for the county-level power distribution network and local microgrids; specifically including the following steps:

[0125] Construct the objective function:

[0126] The following formula is used as the objective function:

[0127] In the formula The set investment cost weighting coefficient; For investment costs, and , For a distributed set of devices, Let i be the capital recovery factor of equipment i and , The annual discount rate is 10%. For the economic service life of device i, Let i be the unit capacity investment cost of equipment. For the additional construction capacity of device i in phase s of scenario s, Sunk costs for reserving capacity per unit of equipment i. For scenario s, stage k represents the capacity reserved for device i. For the collection of distribution network lines, Let be the capital recovery coefficient for line j. Let j be the investment cost per unit length of the line. The new or upgraded length of line j in stage s of scenario s; The set weighting coefficient for operation and maintenance costs; For operation and maintenance costs, and , The total number of time periods within stage k. The duration of a single time step. Let t be the electricity price purchased from the upper-level power grid at time t. Let k be the power purchased by k from the upstream power grid at time t in scenario s. For gas turbine assembly, The unit fuel cost of a gas turbine. For the active power output of device i at time t in phase s of scenario s, The unit maintenance cost of device i. This refers to the collection of all distributed power devices in the system. The system risk cost weighting coefficient is set; For systemic risk costs, and , For the system node set, Let m be the load value of node m. Let be the load reduction amount at time t for node k and node m in scenario s. This is the penalty factor for exceeding the voltage limit per unit. For scenario s, node k and node n at time t, the voltage exceeds the limit and , For reference voltage, This represents the maximum permissible voltage deviation. The set weighting coefficient value for green energy consumption benefits; The revenue from green energy consumption (this item is a negative cost, i.e., a revenue, used to incentivize the full utilization of renewable energy), and , The environmental benefits generated by green electricity per unit. For the actual photovoltaic output of k in phase s at time t, The actual wind power output of k in phase s at time t. The cost of curtailing wind and solar power per unit Let k be the amount of photovoltaic power curtailment at time t in phase s. Let k be the wind power curtailment power at time t in stage s. The duration of each planning phase;

[0128] Construct constraints:

[0129] Active power balance constraints:

[0130] In the formula Let n be the set of all power generation devices connected to node n; Let t be the active power flowing from the upper-level power grid into node n at time t in stage s. Let k be the active power flowing from node n to the upper-level power grid at time t in stage s. Let k be the load active power of node n at time t in phase s of scenario s; Let k be the load reduction amount at node n in phase s at time t. Let n be the set of all charging devices connected to node n. Let be the charging power of energy storage device j at time t in stage k of scenario s;

[0131] Reactive power balance constraints:

[0132] In the formula This is the set of all reactive power compensation devices connected to node n; The reactive power output of device i at time t in phase s of scenario s; Let K be the reactive power flowing from the upper-level power grid into node N at time t in stage s. Let K be the reactive power flowing from node N to the upper-level power grid at time t in stage s. Let K be the reactive power of the load at node n at time t in phase s of scenario s. The set of lines connected to node n; For scenario s, stage k at time t, the path The reactive power consumed;

[0133] Renewable energy output constraints:

[0134] In the formula The rated installed capacity of equipment i is defined as wind turbines and photovoltaics. Let k be the normalized output coefficient of device i at time t in stage s of scenario s;

[0135] Gas turbine operating constraints:

[0136] In the formula To contribute the minimum technical power to the gas turbine; To contribute the most technological expertise to gas turbines; Let k be the start / stop state of the gas turbine in scenario s at time t. If the gas turbine is in the start state in scenario s at time t, then... If, in scenario s, the gas turbine is in a non-starting state at time t, then... ; This represents the maximum ramp power of the gas turbine.

[0137] Energy storage system operating constraints:

[0138] In the formula Let k be the state of charge of the energy stored in scenario s at time t. For energy storage charging efficiency; The discharge efficiency of energy storage; The charging power of energy stored by k in stage s at time t; Let k be the discharge power of the energy stored in k at time t during stage s of scenario s; This refers to the rated capacity of the energy storage. This represents the minimum state of charge allowed for energy storage. This represents the maximum state of charge allowed for energy storage. This represents the maximum charging and discharging power of the energy storage.

[0139] This invention sets the boundary or right-hand side of the following constraints to non-fixed values, i.e., functions related to the planning guidance weight W and the system state; this allows the feasible region of the model to be dynamically adjusted according to different planning objectives, thereby generating solutions that truly meet differentiated needs.

[0140] Voltage quality constraints:

[0141] In the formula Let n be the voltage amplitude at node n at time k in stage s. The set lower voltage limit, and , The specified lower limit of the basic voltage (with a value of...) ), The lower limit of the voltage range at which the set reliability requirement is met (the value is...) ), The set risk weight coefficient; The set upper limit of voltage, and , The specified upper limit of the basic voltage (value is) ), The upper limit of the voltage range required to meet the set reliability requirements (the value is...). );

[0142] Power supply reliability constraints:

[0143] In the formula Total load power; For the maximum allowable expected power shortage, and , Allowed by the established economic orientation Maximum expected power shortage and , Allowed for the set high reliability requirements Maximum expected power shortage and ; The load reduction of node n at time k in phase s; This refers to the risk weighting coefficient.

[0144] This power supply reliability constraint is based on a probabilistic index of expected power shortage, which better reflects the true reliability level of the system. The constraint considers the cumulative power shortage in all scenarios s, all stages k, and all times t, ensuring that the planning scheme meets the reliability requirements throughout the entire life cycle.

[0145] Isolated operation capability constraints:

[0146] In the formula It is the set of all dispatchable distributed power sources within a microgrid; For scenario s at time k The total power of critical loads that must be guaranteed by the microgrid; Based on risk weight coefficient The defined continuous operation time requirement for microgrid islanding, and , The minimum self-balancing time is set. This is the set maximum self-balancing time. Within this constraint... It is no longer a fixed value, but dynamically adjusted according to reliability weights. Furthermore, the constraint requires that for each time t, the future... It achieves self-balancing within a certain time frame, ensuring that the system has the corresponding islanding capability at any time;

[0147] S4. Decompose the model constructed in step S3 to construct a hierarchical distributed solution model; specifically, this includes the following steps:

[0148] The core idea of ​​model decomposition is to decompose the complex global optimization problem according to the physical boundary and the investment entity. Through information decoupling and parallel computing strategies, the computational efficiency is greatly improved while ensuring the accuracy of the solution, and the information privacy of each investment entity is effectively protected.

[0149] The model constructed in step S3 is decomposed into an upper-layer model and a lower-layer model;

[0150] The upper-level model is for distribution networks, and the lower-level model is for microgrids;

[0151] Upper-level model:

[0152] The following formula is used as the objective function of the upper-level model:

[0153] In the formula The investment cost of the distribution network in stage k scenario s; The operating cost of the distribution network in scenario s at stage k; The active power price exchanged between the distribution network and the microgrid at time t in scenario k. This refers to the active power exchanged between the distribution network and the microgrid at time t in scenario k. The distribution network and microgrid exchange reactive power prices at time t in scenario k. This refers to the reactive power exchanged between the distribution network and the microgrid at time t in scenario k. The number of years included in each planning phase;

[0154] Upper-level model constraints:

[0155] Current constraints:

[0156] In the formula Let n be the set of lines ending at node n. This is the set of lines starting at node n; Let be the active power of line l at time t in stage k of scenario s; Let be the current value of line l at time t in stage k of scenario s; Let be the resistance of line l; Let n be the load power of node n at time t in stage k of scenario s; For the set of microgrids connected to node n;

[0157] Voltage coupling constraint:

[0158] In the formula Let be the voltage at connection point m of the microgrid at time t in stage k of scenario s; The voltage at connection point m on the distribution network side at time t in scenario s stage k;

[0159] Line capacity constraints:

[0160] In the formula This represents the maximum transmission capacity of line l. The set line capacity adaptive coefficient;

[0161] Lower-level model:

[0162] The following formula is used as the objective function of the lower-level model:

[0163] In the formula Let m be the investment cost of microgrid m in stage k of scenario s; Let m be the operating cost of microgrid m at time t in stage k of scenario s;

[0164] Lower-level model constraints:

[0165] Power balance constraints within a microgrid:

[0166] In the formula A collection of distributed generation equipment within a microgrid; The active power output of device i at time t in phase k of scenario s; Let m be the energy storage charging power of microgrid m at time t in scenario s phase k. Let be the load reduction amount of microgrid m at time t in phase k of scenario s;

[0167] Equipment operating constraints:

[0168] In the formula Minimum technical contribution to device i; The rated capacity of device i;

[0169] Reliability constraints:

[0170] In the formula The maximum allowable expected power shortage ratio for microgrid m;

[0171] Voltage consistency constraints:

[0172] In the formula Let be the voltage at connection point m on the m-side of the microgrid at time t during scenario s phase k.

[0173] Define the data exchange rules between the upper-level model and the lower-level model:

[0174] The data exchanged between the upper-layer model and the lower-layer model includes power information, voltage information, and price information; among which, power information includes... and Voltage information includes Price information includes and Furthermore, this includes ensuring that the internal cost functions of each microgrid are not disclosed to the distribution network; protecting sensitive information such as the capacity and technical parameters of internal equipment within the microgrid; eliminating the need to share detailed operating strategies within the microgrid; and not exposing commercial information such as the rate of return on investment and internal decision-making mechanisms of each investment entity to external parties.

[0175] S5. Based on the ADMM algorithm, solve the model obtained in step S4 to complete the collaborative planning of the target power system's county-level distribution network and local microgrids, including the following steps:

[0176] This invention is based on the ADMM algorithm and has made some improvements; the ADMM algorithm used in this invention is expressed as follows: st

[0177] In the formula The objective function is the sub-problem of optimizing the upper-level distribution network. The sum of the objective functions of all the lower-level microgrid optimization subproblems; and The coefficient matrix describes the coupling relationships within the system; This represents the decision variable vector of the upper-level distribution network; This is the set of decision variables for all lower-level microgrids; This is the right-hand side term of the linear consistency constraint; This represents a linear consistency constraint;

[0178] The iterative process includes solving the x-subproblem, solving the z-subproblem, and updating the dual variable:

[0179] The solution to the x-subproblem is expressed as:

[0180] In the formula These are the decision variables for the distribution network side obtained in the (k+1)th iteration; Let x represent the augmented Lagrangian function with respect to the variable x; These are the microgrid-side decision variables obtained in the k-th iteration; Let be the dual variable of the k-th iteration;

[0181] The solution to the z-subproblem is expressed as:

[0182] In the formula These are the microgrid-side decision variables obtained in the (k+1)th iteration;

[0183] The update of the dual variable is represented as:

[0184] In the formula For the dual variable updated in the (k+1)th iteration; This is the penalty parameter used in the k-th iteration, and , As the first adjustment factor set, The second adjustment factor is set. For the original residual and , For dual residuals and , This is the 2-norm of a vector, also known as the Euclidean norm. The set balance threshold;

[0185] To address the issues of large probability differences and varying scales of coupling variables in multi-stage stochastic programming, an improved augmented Lagrangian function based on scenario probabilities and adaptive weights is proposed. The augmented Lagrangian function used in this invention is defined as follows:

[0186] In the formula For the k-th iteration, the augmented Lagrangian function is used to coordinate the decomposed distribution network and microgrid subproblems. Represents a weighted matrix The square of the weighted Euclidean norm is used to balance the penalty for violations of constraints on different physical dimensions (such as power and voltage). These are the values ​​of the distribution network-side decision variables obtained in the k-th iteration; The square of the weighted Euclidean norm with respect to the positive definite matrix P is used to measure the change of the distribution network variable x. These are the values ​​of the microgrid-side decision variables obtained in the k-th iteration; The weighted Euclidean norm square with respect to the positive definite matrix Q is used to measure the change of the microgrid variable z; Let be the dynamic regularization parameter at the k-th iteration, and , Initial regularization parameters, It is the attenuation factor, and ; Introduction This makes the degree of satisfaction of coupling constraints in high-probability scenarios contribute more to the overall penalty, guiding the solution to be more inclined towards possible futures; at the same time, the newly added terms As a "soft anchor point," it limits the step size of each iteration when the problem is highly non-convex, effectively preventing iterative oscillations and improving algorithm stability; parameters It can dynamically decay according to the changes in residuals; and the parameters and Instead of being fixed, it is dynamically adjusted based on the original residual and dual residual of each iteration to balance convergence speed and accuracy;

[0187] Map the model obtained in step S4 to the ADMM algorithm framework: map the objective function of the upper-level model to... The objective function of the lower-level model is corresponding to Linear consistency constraints include power balance constraints, voltage consistency constraints, and data exchange rules between the upper and lower level models; These are the decision variables for the higher-level model. For the decision variables of the lower-level model, c includes the power exchange boundary and voltage reference value;

[0188] At the beginning of each planning cycle, the ADMM algorithm initializes x, z, and y using the optimal solution from the previous stage; the state variables between planning cycles are passed through Markov chains.

[0189] For the lower-level model, an asynchronous parallel solution strategy is adopted: each microgrid independently solves its own z-subproblem, without considering the iteration completion status of other microgrids; the upper-level model, after collecting a set number of results from the lower-level model, solves its own x-subproblem to reduce synchronization waiting time; the asynchronous update rule is expressed as: for microgrid m, the corresponding z-subproblem is solved using the latest and most usable y and x; the upper-level model, after receiving .... After obtaining the results of the lower-level model, solve its own x-subproblems;

[0190] The standardized residuals are set as follows:

[0191] In the formula To standardize the original residual, it is used to measure the degree of violation of coupling constraints by the current iterative solution. The smaller the value, the better the physical constraints such as the overall power balance and voltage consistency of the system are satisfied. To standardize the dual residuals, which measure the drastic change in the dual variables, a smaller value indicates a more stable consensus in the distributed optimization process.

[0192] Set the maximum number of iterations to ;

[0193] Set the iteration termination condition:

[0194] Condition 1: ;in The first tolerance threshold is set;

[0195] Condition 2: ;in This is the set second tolerance threshold;

[0196] Condition 3: The number of iterations has been reached. ;

[0197] The ADMM algorithm used stops iterating when any one of condition 1, condition 2, or condition 3 is met.

[0198] The following steps can be used to solve the problem:

[0199] (1) Initialization: Model parameter initialization;

[0200] (2) The lower-level model is solved in parallel: each microgrid independently optimizes its own scheme;

[0201] (3) Solving the upper-level model: Each distribution network optimizes its own scheme;

[0202] (4) Coordination variable update;

[0203] (5) Adjust algorithm parameters and optimize model performance;

[0204] (6) Convergence judgment; output the final planning scheme.

[0205] This invention constructs an innovative technical system that fully covers the entire process of "modeling-decomposition-solution". Firstly, it effectively handles multiple uncertainties such as load, renewable energy, and cost during the planning period through a multi-stage stochastic programming framework and improved scenario tree generation technology. Secondly, it designs a differentiated objective function and adaptive constraint system to dynamically integrate the heterogeneous demands of county-level regions into the model, achieving a flexible trade-off between economic efficiency, reliability, and green goals. Addressing the solution challenges posed by the high dimensionality and nonlinearity of the model, a hierarchical distributed solution framework is proposed, decomposing the complex problem into distribution network and multiple microgrid sub-problems through physical and information decoupling. Finally, an improved adaptive alternating direction multiplier method (ADMM) is adopted, introducing scenario probability weighting, dynamic regularization, and asynchronous parallel mechanisms to significantly improve the convergence speed and stability of the algorithm while ensuring data privacy for all parties. This invention achieves efficient collaborative planning that balances regional differences and protects privacy information under uncertain environments, providing a systematic solution for the precise investment and optimized operation of county-level distribution networks and microgrids.

[0206] like Figure 2 The diagram shows the functional modules of the system of the present invention: The system disclosed in this invention for implementing the collaborative planning method of the county-level power distribution network and local microgrids of the power system includes a data acquisition module, a scenario construction module, a model construction module, a model decomposition module, and a collaborative planning module; the data acquisition module, scenario construction module, model construction module, model decomposition module, and collaborative planning module are connected in series; the data acquisition module is used to acquire data information of the county-level power distribution network and local microgrids of the target power system and upload the data information to the scenario construction module; the scenario construction module is used to construct the county-level power distribution network of the target power system based on the received data information and the acquired data information. The system coordinates the planning and operation of the power system's county-level distribution network and local microgrids, and uploads the data to the model building module. The model building module, based on the received data and the constructed operation scenario, builds a collaborative planning model for the power system's county-level distribution network and local microgrids, and uploads the data to the model decomposition module. The model decomposition module, based on the received data, decomposes the constructed model to build a hierarchical distributed solution model, and uploads the data to the collaborative planning module. The collaborative planning module, based on the received data and using the ADMM algorithm, solves the obtained model to complete the collaborative planning of the target power system's county-level distribution network and local microgrids.

Claims

1. A collaborative planning method for county-level power distribution networks and local microgrids in a power system, comprising the following steps: S1. Obtain data information on the county-level distribution network and local microgrids of the target power system; S2. Based on the data obtained in step S1, construct a collaborative planning and operation scenario for the county-level power distribution network and local microgrids of the target power system; S3. Based on the operational scenario constructed in step S2, construct a collaborative planning model for the county-level power distribution network and local microgrids in the power system; S4. Decompose the model constructed in step S3 to construct a hierarchical distributed solution model; S5. Based on the ADMM algorithm, solve the model obtained in step S4 to complete the collaborative planning of the target power system county distribution network and local microgrid.

2. The collaborative planning method for county-level power distribution networks and local microgrids according to claim 1, characterized in that... Step S2, which involves constructing a collaborative planning and operation scenario for the target power system's county-level distribution network and local microgrids based on the data obtained in step S1, specifically includes the following steps: Planning cycle The decision is divided into K consecutive decision-making stages; the starting time of each stage k is... ; Based on historical data and a pre-trained large model, several random paths are generated for each uncertainty factor, and each path represents a possible future evolution trajectory. The generated paths are used to construct an initial scene tree T; each node n in the initial scene tree T represents a possible state at the corresponding time point; A path from the root node to a leaf node of the initial scene tree is called a scene s, and the corresponding probability is . ; The initial scene tree is reduced using the following steps: Calculate any two scenarios s and s using the following formula. Distance between : In the formula Let s be the vector of uncertainty variables in scenario s at stage k, including load. Photovoltaic cost Wind power costs Light intensity and wind speed ; It is the 2-norm of the vector; Let be the overall volatility index for scenario s, and , Let be the standard deviation of each uncertainty variable in the k-th stage; Let be the average correlation coefficient among different uncertainty variables in scenario s; The set fluctuation weight; The relevant weights are set; Calculate all possible pairs of scenes s and Distance between And construct the corresponding distance matrix. for ; Hierarchical clustering algorithm is used for scene clustering; For each class The scene with the smallest average distance from other scenes was selected as the representative scene. ; The probability representing the scenario Set as ; After the above steps, the initial n scenarios are reduced to M representative scenarios and their corresponding probabilities. ; Complete the construction of a collaborative planning and operation scenario for the county-level distribution network and local microgrids of the target power system.

3. The collaborative planning method for county-level power distribution networks and local microgrids according to claim 2, characterized in that... Step S3, which involves constructing a collaborative planning model for the county-level power distribution network and local microgrids based on the operational scenario established in step S2, specifically includes the following steps: Construct the objective function: The following formula is used as the objective function: In the formula The set investment cost weighting coefficient; For investment costs, and , For a distributed set of devices, Let i be the capital recovery factor of equipment i and , The annual discount rate is 10%. For the economic service life of device i, Let i be the unit capacity investment cost of equipment. For the additional construction capacity of device i in phase s of scenario s, Sunk costs for reserving capacity per unit of equipment i. For scenario s, stage k represents the capacity reserved for device i. For the collection of distribution network lines, Let be the capital recovery coefficient for line j. Let j be the investment cost per unit length of the line. The new or upgraded length of line j in stage s of scenario s; The set weighting coefficient for operation and maintenance costs; For operation and maintenance costs, and , The total number of time periods within stage k. The duration of a single time step. Let t be the electricity price purchased from the upper-level power grid at time t. Let k be the power purchased by k from the upstream power grid at time t in scenario s. For gas turbine assembly, The unit fuel cost of a gas turbine. For the active power output of device i at time t in phase s of scenario s, The unit maintenance cost of device i. This refers to the collection of all distributed power devices in the system. The system risk cost weighting coefficient is set; For systemic risk costs, and , For the system node set, Let m be the load value of node m. Let be the load reduction amount at time t for node k and node m in scenario s. This is the penalty factor for exceeding the voltage limit per unit. For scenario s, node k and node n at time t, the voltage exceeds the limit and , For reference voltage, This represents the maximum permissible voltage deviation. The set weighting coefficient value for green energy consumption benefits; For the benefit of green energy consumption, and , The environmental benefits generated by green electricity per unit. For the actual photovoltaic output of k in phase s at time t, The actual wind power output of k in phase s at time t. The cost of curtailing wind and solar power per unit Let k be the amount of photovoltaic power curtailment at time t in phase s. Let k be the wind power curtailment power at time t in stage s. The duration of each planning phase; Construct constraints: Active power balance constraints: In the formula Let n be the set of all power generation devices connected to node n; Let t be the active power flowing from the upper-level power grid into node n at time t in stage s. Let k be the active power flowing from node n to the upper-level power grid at time t in stage s. Let k be the load active power of node n at time t in phase s of scenario s; Let k be the load reduction amount at node n in phase s at time t. Let n be the set of all charging devices connected to node n. Let be the charging power of energy storage device j at time t in stage k of scenario s; Reactive power balance constraints: In the formula This is the set of all reactive power compensation devices connected to node n; The reactive power output of device i at time t in phase s of scenario s; Let K be the reactive power flowing from the upper-level power grid into node N at time t in stage s. Let K be the reactive power flowing from node N to the upper-level power grid at time t in stage s. Let K be the reactive power of the load at node n at time t in phase s of scenario s. The set of lines connected to node n; For scenario s, stage k at time t, the path The reactive power consumed; Renewable energy output constraints: In the formula The rated installed capacity of equipment i is defined as wind turbines and photovoltaics. Let k be the normalized output coefficient of device i at time t in stage s of scenario s; Gas turbine operating constraints: In the formula To contribute the minimum technical power to the gas turbine; To contribute the most technological expertise to gas turbines; Let k be the start / stop state of the gas turbine in scenario s at time t. If the gas turbine is in the start state in scenario s at time t, then... If, in scenario s, the gas turbine is in a non-starting state at time t, then... ; This represents the maximum ramp power of the gas turbine. Energy storage system operating constraints: In the formula Let k be the state of charge of the energy stored in scenario s at time t. For energy storage charging efficiency; The discharge efficiency of energy storage; The charging power of energy stored by k in stage s at time t; Let k be the discharge power of the energy stored in k at time t during stage s of scenario s; This refers to the rated capacity of the energy storage. This represents the minimum state of charge allowed for energy storage. This represents the maximum state of charge allowed for energy storage. This represents the maximum charging and discharging power of the energy storage. Voltage quality constraints: In the formula Let n be the voltage amplitude at node n at time k in stage s. The set lower voltage limit, and , The specified lower limit of the basic voltage. To ensure that the set reliability requirements reach the lower limit of the voltage range, The set risk weight coefficient; The set upper limit of voltage, and , The upper limit of the basic voltage is specified. The upper limit of the voltage range required to meet the set reliability requirements; Power supply reliability constraints: In the formula Total load power; For the maximum allowable expected power shortage, and , Allowed by the established economic orientation Maximum expected power shortage and , Allowed for the set high reliability requirements Maximum expected power shortage and ; The load reduction of node n at time k in phase s; This refers to the risk weighting coefficient. Isolated operation capability constraints: In the formula It is the set of all dispatchable distributed power sources within a microgrid; For scenario s at time k The total power of critical loads that must be guaranteed by the microgrid; Based on risk weight coefficient The defined continuous operation time requirement for microgrid islanding, and , The minimum self-balancing time is set. The set maximum self-balancing time.

4. The collaborative planning method for county-level power distribution networks and local microgrids according to claim 3, characterized in that... Step S4, which involves decomposing the model constructed in step S3 to build a hierarchical distributed solution model, specifically includes the following steps: The model constructed in step S3 is decomposed into an upper-layer model and a lower-layer model; The upper-level model is for distribution networks, and the lower-level model is for microgrids; Upper-level model: The following formula is used as the objective function of the upper-level model: In the formula The investment cost of the distribution network in stage k scenario s; The operating cost of the distribution network in scenario s at stage k; The active power price exchanged between the distribution network and the microgrid at time t in scenario k. This refers to the active power exchanged between the distribution network and the microgrid at time t in scenario k. The distribution network and microgrid exchange reactive power prices at time t in scenario k. This refers to the reactive power exchanged between the distribution network and the microgrid at time t in scenario k. The number of years included in each planning phase; Upper-level model constraints: Current constraints: In the formula Let n be the set of lines ending at node n. This is the set of lines starting at node n; Let be the active power of line l at time t in stage k of scenario s; Let be the current value of line l at time t in stage k of scenario s; Let be the resistance of line l; Let n be the load power of node n at time t in stage k of scenario s; For the set of microgrids connected to node n; Voltage coupling constraint: In the formula Let be the voltage at connection point m of the microgrid at time t in stage k of scenario s; The voltage at connection point m on the distribution network side at time t in scenario s stage k; Line capacity constraints: In the formula This represents the maximum transmission capacity of line l. The set line capacity adaptive coefficient; Lower-level model: The following formula is used as the objective function of the lower-level model: In the formula Let m be the investment cost of microgrid m in stage k of scenario s; Let m be the operating cost of microgrid m at time t in stage k of scenario s; Lower-level model constraints: Power balance constraints within a microgrid: In the formula A collection of distributed generation equipment within a microgrid; The active power output of device i at time t in phase k of scenario s; Let m be the energy storage charging power of microgrid m at time t in scenario s phase k. Let be the load reduction amount of microgrid m at time t in phase k of scenario s; Equipment operating constraints: In the formula Minimum technical contribution to device i; The rated capacity of device i; Reliability constraints: In the formula The maximum allowable expected power shortage ratio for microgrid m; Voltage consistency constraints: In the formula Let be the voltage at connection point m on the m-side of the microgrid at time t during scenario s phase k. Define the data exchange rules between the upper-level model and the lower-level model: The data exchanged between the upper-layer model and the lower-layer model includes power information, voltage information, and price information; among which, power information includes... and Voltage information includes Price information includes and .

5. The collaborative planning method for county-level power distribution networks and local microgrids according to claim 4, characterized in that... Step S5, which involves solving the model obtained in step S4 using the ADMM algorithm, includes the following steps: The ADMM algorithm used is represented as follows: s.t. In the formula The objective function is the sub-problem of optimizing the upper-level distribution network. The sum of the objective functions of all the lower-level microgrid optimization subproblems; and The coefficient matrix describes the coupling relationships within the system; This represents the decision variable vector of the upper-level distribution network; This is the set of decision variables for all lower-level microgrids; This is the right-hand side term of the linear consistency constraint; This represents a linear consistency constraint; The iterative process includes solving the x-subproblem, solving the z-subproblem, and updating the dual variable: The solution to the x-subproblem is expressed as: In the formula These are the decision variables for the distribution network side obtained in the (k+1)th iteration; Let x represent the augmented Lagrangian function with respect to the variable x; These are the microgrid-side decision variables obtained in the k-th iteration; Let be the dual variable of the k-th iteration; The solution to the z-subproblem is expressed as: In the formula These are the microgrid-side decision variables obtained in the (k+1)th iteration; The update of the dual variable is represented as: In the formula For the dual variable updated in the (k+1)th iteration; This is the penalty parameter used in the k-th iteration, and , As the first adjustment factor set, The second adjustment factor is set. For the original residual and , For dual residuals and , Let be the 2-norm of the vector. The set balance threshold; The augmented Lagrangian function is defined as follows: In the formula Let be the augmented Lagrangian function for the k-th iteration; Represents a weighted matrix The square of the weighted Euclidean norm; These are the values ​​of the distribution network-side decision variables obtained in the k-th iteration; Let be the square of the weighted Euclidean norm with respect to the positive definite matrix P; These are the values ​​of the microgrid-side decision variables obtained in the k-th iteration; Let be the weighted Euclidean norm squared with respect to the positive definite matrix Q; Let be the dynamic regularization parameter at the k-th iteration, and , These are the initial regularization parameters. It is the attenuation factor and ; Map the model obtained in step S4 to the ADMM algorithm framework: map the objective function of the upper-level model to... The objective function of the lower-level model is corresponding to Linear consistency constraints include power balance constraints, voltage consistency constraints, and data exchange rules between the upper and lower level models; These are the decision variables for the higher-level model. For the decision variables of the lower-level model, c includes the power exchange boundary and voltage reference value; At the beginning of each planning cycle, the ADMM algorithm initializes x, z, and y using the optimal solution from the previous stage; the state variables between planning cycles are passed through Markov chains. For the lower-level model, an asynchronous parallel solution strategy is adopted: each microgrid independently solves its own z-subproblem, without considering the iteration completion status of other microgrids; the upper-level model, after collecting a set number of results from the lower-level model, solves its own x-subproblem to reduce synchronization waiting time; the asynchronous update rule is expressed as: for microgrid m, the corresponding z-subproblem is solved using the latest and most usable y and x; the upper-level model, after receiving .... After obtaining the results of the lower-level model, solve its own x-subproblems; The standardized residuals are set as follows: In the formula To standardize the original residuals; To standardize the dual residuals; Set the maximum number of iterations to ; Set the iteration termination condition: Condition 1: ;in The first tolerance threshold is set; Condition 2: ;in This is the set second tolerance threshold; Condition 3: The number of iterations has been reached. ; The ADMM algorithm used stops iterating when any one of condition 1, condition 2, or condition 3 is met.

6. The collaborative planning method for county-level power distribution networks and local microgrids according to claim 5, characterized in that... The solution described in step S5 specifically includes the following steps: (1) Initialization: Model parameter initialization; (2) The lower-level model is solved in parallel: each microgrid independently optimizes its own scheme; (3) Solving the upper-level model: Each distribution network optimizes its own scheme; (4) Coordination variable update; (5) Adjust algorithm parameters and optimize model performance; (6) Convergence judgment; output the final planning scheme.

7. A system for implementing the collaborative planning method for county-level power distribution networks and local microgrids as described in any one of claims 1 to 6, characterized in that... It includes a data acquisition module, a scenario construction module, a model construction module, a model decomposition module, and a collaborative planning module; these modules are connected in series. The data acquisition module acquires data information of the county-level distribution network and local microgrids of the target power system and uploads the data information to the scenario construction module. The scenario construction module constructs a collaborative planning and operation scenario of the county-level distribution network and local microgrids of the target power system based on the received and acquired data information, and uploads the data information to the model construction module. The model building module is used to construct a collaborative planning model of the county-level power distribution network and local microgrids based on the received data and the constructed operating scenario, and upload the data to the model decomposition module. The model decomposition module is used to decompose the constructed model based on the received data information to build a hierarchical distributed solution model, and upload the data information to the collaborative planning module. The collaborative planning module is used to solve the obtained model based on the received data information and the ADMM algorithm to complete the collaborative planning of the target power system county distribution network and local microgrids.

Images