Method, system and device for modeling and economic dispatch of grid-scale energy storage and medium

Abstract

The application discloses a network-constructed energy storage modeling and economic dispatching method, system, device and medium, and relates to the technical field of energy storage control of power systems; including: arranging intelligent power exchange base stations in an energy storage subnetwork according to hexagon vertex coordinates to form a hexagon unit; generating a topology connection matrix, initial node voltage and initial angular frequency, a main ring network admittance matrix and an auxiliary ring network admittance matrix according to the hexagon unit and a double ring topology; performing quantum field cooperative control by constructing a quantum state Hamiltonian and solving a Schrodinger equation to generate a cooperative control instruction and an intelligent power exchange base station working mode vector; then performing asymmetric dynamic impedance remodeling, calculating dynamic admittance of each node, and converting the dynamic admittance into a time-varying impedance matrix; and generating a game equilibrium cost by solving a Nash equilibrium through quantum annealing. The application solves the problems of insufficient transient response capability and low optimization scheduling efficiency in the prior art.

Description

Technical Field

[0001] This invention relates to the field of power system energy storage control technology, and in particular to a modeling and economic dispatch method, system, equipment and medium for grid-connected energy storage. Background Technology

[0002] With the continuous increase in renewable energy penetration, grid-based energy storage systems have become one of the key technologies supporting the stable operation of new power systems. Currently, under a high proportion of renewable energy grid connection, the system faces prominent problems such as delayed transient response, difficulty in suppressing circulating currents, and the challenge of coordinating economic efficiency with stability. Especially during peak photovoltaic power generation periods at midday or under extreme weather conditions, the power grid places higher demands on the rapid adjustment capabilities and economic operating efficiency of energy storage systems.

[0003] Existing grid-based energy storage control technologies mainly include virtual synchronous machine (VSG) control and fixed impedance-based collaborative strategies. These methods improve the system's inertia response and voltage support capabilities to some extent, but in complex scenarios with multiple subgrids interconnected and dynamically changing topologies, it remains difficult to balance the speed, robustness, and economic requirements of multi-subgrid collaboration. Traditional optimization methods, such as mixed-integer second-order cone programming (MI-SOCP), can achieve economical dispatch, but their long solution time makes it difficult to meet the collaborative requirements of millisecond-level control and minute-level optimization.

[0004] Therefore, existing technologies mainly suffer from the following problems: First, insufficient transient response capability. Traditional control methods have long inertial response times, low voltage recovery rates, and limited fault ride-through success rates, making it difficult to provide millisecond-level rapid dynamic support in scenarios with a high proportion of renewable energy grid integration. Second, low efficiency in optimized scheduling. Existing economic scheduling methods are time-consuming to solve, failing to achieve rapid coordination between stability control and economic operation, resulting in electricity purchase costs and operational risks that are difficult to achieve ideal levels. Summary of the Invention

[0005] This invention provides a modeling and economic dispatch method, system, equipment, and medium for grid-connected energy storage, which addresses the problems of insufficient transient response capability and low efficiency of optimized dispatch in existing technologies.

[0006] In view of this, the first aspect of the present invention provides a modeling and economic dispatch method for grid-connected energy storage, the method comprising:

[0007] Within each energy storage subnet, intelligent power switching base stations are deployed according to the hexagonal vertex coordinates to form hexagonal units; a double ring topology is constructed between each subnet to connect them; based on the hexagonal units and the double ring topology, a topology connection matrix, initial node voltage and initial angular frequency, main ring network admittance matrix and auxiliary ring network admittance matrix are generated.

[0008] Based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, quantum field cooperative control is performed by constructing quantum state Hamiltonians and solving the Schrödinger equation to generate cooperative control commands and intelligent power exchange base station operating mode vectors.

[0009] Based on the cooperative control command and the working mode vector of the intelligent power switching base station, asymmetric dynamic impedance reshaping is performed, the dynamic admittance of each node is calculated, and the dynamic admittance is converted into a time-varying impedance matrix.

[0010] Based on the time-varying impedance matrix and combined with the electricity price signal, master-slave game optimization is performed, and the Nash equilibrium is solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost.

[0011] Optionally, within each energy storage subnet, intelligent power switching base stations are deployed according to the hexagonal vertex coordinates to form hexagonal units; a dual-ring topology is constructed between each subnet to connect them; based on the hexagonal units and the dual-ring topology, a topology connection matrix, initial node voltages and initial angular frequencies, a main ring network admittance matrix, and an auxiliary ring network admittance matrix are generated, including:

[0012] Within each energy storage subnetwork, each smart power exchange base station is deployed according to the vertex coordinates of a regular hexagon, so that each base station constitutes a hexagonal unit.

[0013] Between each energy storage subnetwork, a main ring network and an auxiliary ring network are constructed. The main ring network connects the central nodes of each subnetwork, and the auxiliary ring network forms virtual connections between vertex nodes through Kron reduction. The admittance matrix of the main ring network is established based on the line impedance parameters, and the admittance matrix of the auxiliary ring network is established based on Kron reduction.

[0014] Based on the actual connection relationships between each node, a topology connection matrix is ​​generated, where the positions of directly connected nodes are marked as connected states, and those not connected are marked as non-connected states.

[0015] Based on the admittance matrix of the main ring network, the admittance matrix of the auxiliary ring network, and the injected current, the initial node voltage is obtained by solving the node voltage equation.

[0016] Optionally, the step of performing quantum field cooperative control based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, by constructing a quantum state Hamiltonian and solving the Schrödinger equation, to generate cooperative control commands and a smart power exchange base station operating mode vector, includes:

[0017] Based on the initial node voltage, the initial angular frequency, the main ring network admittance matrix, and the auxiliary ring network admittance matrix, a quantum state Hamiltonian is constructed, which includes a kinetic energy term, a voltage potential energy term, and a frequency coupling potential energy term.

[0018] The initial quantum state amplitude of each subnet is determined based on the topological connection matrix, and the Schrödinger equation is solved by combining the Hamiltonian to obtain the evolution result of the joint quantum state of multiple subnets over time.

[0019] Based on the evolution of the multi-subnet joint quantum state over time, the voltage correction of each node is calculated by quantum expectation value inversion, and cooperative control commands are generated.

[0020] Based on the evolution of the joint quantum state of the multi-subnet over time, the quantum tunneling probability of each subnet is calculated. When the quantum tunneling probability is less than a preset threshold, the working mode of the intelligent power switching base station in that subnet is switched, and an intelligent power switching base station working mode vector is generated.

[0021] Optionally, the expression for the quantum state Hamiltonian is:

[0022] ;

[0023] In the formula, The Hamiltonian of the quantum state is... For nodes The generalized momentum operator, For node equivalent quality, Let be the potential energy function.

[0024] Optionally, the step of performing asymmetric dynamic impedance reshaping based on the cooperative control command and the intelligent power switching base station operating mode vector, calculating the dynamic admittance of each node, and converting the dynamic admittance into a time-varying impedance matrix includes:

[0025] The voltage correction amount of each node is extracted from the cooperative control command, and the admittance gain coefficient of each node is calculated based on the voltage correction amount.

[0026] Based on the working mode vector of the intelligent power switching base station, determine the asymmetric delay factor and master-slave weight coefficient of each node;

[0027] The dynamic admittance of each node is calculated based on the admittance gain coefficient, the asymmetric delay factor, and the preset lead time constant and lag time constant.

[0028] The dynamic admittance of each node is converted into dynamic impedance, and then weighted and synthesized using the master-slave weight coefficients of each node to obtain the time-varying impedance matrix.

[0029] Optionally, the expression for calculating the dynamic admittance is:

[0030] ;

[0031] In the formula, Let k be the dynamic admittance of node k. The admittance gain coefficient, The lead time constant, The time lag constant is It is an asymmetric delay factor. It is a Laplace complex variable.

[0032] Optionally, the step of performing master-slave game optimization based on the time-varying impedance matrix and combined with the electricity price signal, and solving the Nash equilibrium through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost includes:

[0033] The power transmission loss coefficient of each node is calculated based on the time-varying impedance matrix, and an upper-level global optimization model is constructed in conjunction with the electricity price signal. With the goal of minimizing the electricity purchase cost and risk cost, the total scheduling power of each node is obtained.

[0034] The positive-sequence impedance and negative-sequence impedance of each node are calculated based on the time-varying impedance matrix, thereby determining the charging and discharging efficiency of each energy storage. Combined with the total scheduling power and the charging and discharging efficiency of each energy storage, a lower-level local optimization model is constructed to determine the charging and discharging state variables of each energy storage with the goal of maximizing the revenue of each node.

[0035] The upper-level global optimization model and the lower-level local optimization model are transformed into the Ising model. The objective function of the Ising model includes a quadratic term and a linear term composed of the charging and discharging state variables.

[0036] The ground state of the Ising model is solved by quantum annealing to obtain the charging and discharging state sequence of each energy storage unit. The actual charging and discharging power of each energy storage unit is calculated based on the charging and discharging state sequence to generate an energy storage power scheduling plan.

[0037] Based on the energy storage power dispatch plan, the electricity purchase cost, risk cost, and control deviation penalty are calculated and summed to obtain the game equilibrium cost.

[0038] A second aspect of the present invention provides a modeling and economic dispatch system for grid-connected energy storage, the system comprising:

[0039] The first computing unit is used to deploy smart power switching base stations according to the hexagonal vertex coordinates within each energy storage subnet to form a hexagonal unit; to construct a double ring topology between each subnet to connect the subnets; and to generate a topology connection matrix, initial node voltage and initial angular frequency, main ring network admittance matrix and auxiliary ring network admittance matrix based on the hexagonal unit and the double ring topology.

[0040] The second computing unit is used to perform quantum field cooperative control by constructing quantum state Hamiltonians and solving the Schrödinger equation based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, so as to generate cooperative control commands and intelligent power exchange base station operating mode vectors.

[0041] The third calculation unit is used to perform asymmetric dynamic impedance reshaping based on the cooperative control command and the working mode vector of the intelligent power exchange base station, calculate the dynamic admittance of each node, and convert the dynamic admittance into a time-varying impedance matrix.

[0042] The fourth calculation unit is used to perform master-slave game optimization based on the time-varying impedance matrix and the electricity price signal, solve the Nash equilibrium through quantum annealing, and generate an energy storage power scheduling plan and game equilibrium cost.

[0043] A third aspect of the present invention provides a modeling and economic dispatching device for grid-connected energy storage, the device comprising a processor and a memory:

[0044] The memory is used to store program code and transmit the program code to the processor;

[0045] The processor is used to execute the steps of the modeling and economic dispatch method for grid-connected energy storage as described in the first aspect above, according to the instructions in the program code.

[0046] A fourth aspect of the present invention provides a computer-readable storage medium for storing program code for executing the modeling and economic dispatch method for grid-based energy storage described in the first aspect above.

[0047] As can be seen from the above technical solutions, the present invention has the following advantages:

[0048] This invention provides a modeling and economic dispatch method for grid-based energy storage. First, within each energy storage subgrid, intelligent power exchange base stations are deployed according to the vertex coordinates of a regular hexagon to form hexagonal units. A dual-ring topology consisting of a main ring network and an auxiliary ring network is constructed between each subgrid. Based on this, a topology connection matrix, initial node voltages and angular frequencies, and admittance matrices for the main and auxiliary ring networks are generated. The hexagonal symmetric structure naturally cancels circulating currents and improves fault tolerance. The main ring network transmits active power, and the auxiliary ring network constructs virtual inertia channels through Kron reduction, laying a regular electrical foundation for subsequent rapid coordinated control. Then, based on the above topology and admittance matrix, a quantum state Hamiltonian containing kinetic energy, voltage potential energy, and frequency coupling potential energy is constructed. The Schrödinger equation is solved to obtain the multi-subgrid joint quantum state evolution results, and then quantum field coordinated control is executed to generate coordinated control commands and intelligent power exchange base station operating mode vectors. By actively suppressing circulating current using the quantum tunneling effect and rapidly generating voltage correction values ​​through quantum expectation value inversion, the inertia response time is significantly shortened, voltage recovery rate and fault ride-through success rate are improved, and transient response capability is greatly enhanced. Secondly, based on the cooperative control commands and operating mode vectors, the admittance gain coefficient, asymmetric delay factor, and master-slave weight coefficient of each node are calculated, thereby obtaining dynamic admittance and converting it into a time-varying impedance matrix. This process achieves asymmetric dynamic impedance reshaping, enabling impedance parameters to adaptively adjust in real time according to system state, further suppressing circulating current and enhancing dynamic support, effectively improving system robustness. Finally, based on the time-varying impedance matrix and electricity price signal, an upper-level global optimization model (minimizing electricity purchase cost and conditional risk value) and a lower-level local optimization model (maximizing the revenue of each node) are constructed. The two-layer game is transformed into an Ising model, and the Nash equilibrium is quickly solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost. Quantum annealing optimization significantly shortens the solution time, achieving synergy between millisecond-level control and minute-level optimization, effectively reducing electricity purchase cost and operational risk. In summary, this invention solves the problems of insufficient transient response capability and low optimization scheduling efficiency in the prior art. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart illustrating a modeling and economic dispatch method for grid-connected energy storage provided in an embodiment of the present invention;

[0051] Figure 2A schematic diagram of the honeycomb hexagonal cell and double ring network topology provided in an embodiment of the present invention;

[0052] Figure 3 This is a schematic diagram of a grid-based energy storage modeling and economic dispatch system provided in an embodiment of the present invention. Detailed Implementation

[0053] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0054] Please see Figure 1 The present invention provides a method for modeling and economically dispatching grid-connected energy storage, comprising:

[0055] Step 101: Within each energy storage subnet, deploy intelligent power switching base stations according to the hexagonal vertex coordinates to form hexagonal units; construct a double ring topology between subnets to connect them; generate a topology connection matrix, initial node voltage and initial angular frequency, main ring network admittance matrix and auxiliary ring network admittance matrix based on the hexagonal units and the double ring topology.

[0056] Understandably, the first step is to deploy smart power switching base stations within each energy storage subgrid according to the vertex coordinates of a regular hexagon, forming a hexagonal unit with a symmetrical structure, such as... Figure 2 As shown, its geometric properties naturally cancel circulating currents and improve fault tolerance. Then, a double-ring topology consisting of a main ring network and an auxiliary ring network is constructed between each subnet. The main ring network connects the central node for transmitting active power, and the auxiliary ring network forms virtual connections between vertex nodes through Kron reduction to transmit virtual inertia signals. Finally, based on the above topology, the topology connection matrix, initial node voltages and angular frequencies, the admittance matrix of the main ring network, and the admittance matrix of the auxiliary ring network are generated, providing a regular electrical model and steady-state initial values ​​for subsequent quantum field cooperative control and asymmetric dynamic impedance reshaping.

[0057] Step 102: Based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, quantum field cooperative control is performed by constructing quantum state Hamiltonians and solving the Schrödinger equation to generate cooperative control commands and intelligent power exchange base station operating mode vectors.

[0058] Understandably, based on the generated topology connection matrix, initial node voltages and angular frequencies, and admittance matrices of the main and auxiliary ring networks, a quantum state Hamiltonian containing kinetic energy, voltage potential energy, and frequency coupling potential energy terms is constructed. The Schrödinger equation is then solved to obtain the evolution of the joint quantum state of the multi-subnet network over time, thereby executing quantum field cooperative control. On one hand, the voltage correction of each node is calculated using quantum expectation value inversion as a cooperative control command. On the other hand, the operating mode of the intelligent power exchange base station is switched according to the quantum tunneling probability determination result, ultimately generating cooperative control commands and operating mode vectors. The key role of this step is to dynamically map the power grid as a quantum system, actively suppressing circulating currents and accelerating transient response through the quantum tunneling effect, and achieving rapid cooperation among multiple subnets through quantum state evolution, thereby significantly improving the system's transient response capability.

[0059] Step 103: Based on the cooperative control command and the working mode vector of the intelligent power switching base station, perform asymmetric dynamic impedance reshaping, calculate the dynamic admittance of each node, and convert the dynamic admittance into a time-varying impedance matrix;

[0060] Understandably, based on the voltage correction provided by the coordinated control commands and the asymmetric delay factor and master-slave weight coefficient determined by the intelligent power exchange base station's operating mode vector, the dynamic admittance of each node is calculated and converted into a time-varying impedance matrix. The key to this process lies in "asymmetric dynamic impedance reshaping," which enables the impedance parameters to adaptively adjust according to the system's real-time operating status (such as voltage deviation, operating mode, and master-slave roles). This effectively suppresses circulating current, improves transient response, and provides an accurate impedance model that changes with operating conditions for subsequent economic dispatch.

[0061] Step 104: Based on the time-varying impedance matrix and combined with the electricity price signal, perform master-slave game optimization, solve the Nash equilibrium through quantum annealing, and generate the energy storage power scheduling plan and game equilibrium cost.

[0062] Understandably, based on the time-varying impedance matrix obtained in step 103 and the real-time electricity price signal, a master-slave game optimization is performed: the upper layer determines the total dispatch power of each node with the goal of minimizing electricity purchase cost and risk cost, while the lower layer determines the charging and discharging state variables of each energy storage unit with the goal of maximizing the revenue of each node. The two-layer optimization problem is transformed into the Ising model, and the Nash equilibrium is quickly solved through quantum annealing, ultimately generating the energy storage power dispatch plan and the game equilibrium cost. The key role of this step is to embed the impact of dynamic impedance on power loss and charging and discharging efficiency into the economic dispatch model, and to significantly shorten the solution time by utilizing the efficient parallel search capability of quantum annealing, thereby achieving rapid coordination between stability control and economic operation, and effectively solving the problem of low efficiency in optimization dispatch.

[0063] In one embodiment, step 101 includes:

[0064] Step 1011: Within each energy storage subnetwork, each smart power exchange base station is deployed according to the vertex coordinates of a regular hexagon, so that each base station constitutes a hexagonal unit.

[0065] Specifically, in the initialization of the cellular topology, each energy storage subnetwork forms a hexagonal unit through a Smart Power Exchange Base Station (SPIES), such as... Figure 2 As shown. Each energy storage subgrid is connected via a dual-ring topology; and the following parameters are input: grid topology parameters (node ​​coordinates) impedance parameters Energy storage unit configuration (capacity) ); New energy power output forecast data .

[0066] The processing flow is as follows:

[0067] Constructing hexagonal honeycomb cells: Each SPIES node is positioned according to hexagonal coordinates, satisfying:

[0068] ;

[0069]

[0070] Geometric interpretation: This formula generates the vertex coordinates and spacing between adjacent nodes of a regular hexagon in a two-dimensional plane. .

[0071] In this embodiment, the central node of the hexagonal unit is used to deploy a grid-type energy storage converter, serving as the core of the subgrid control; the vertex nodes are used to connect to new energy units or loads such as photovoltaic and wind power, and are accessed through power electronic interfaces. Furthermore, the edges are connected: each edge of the hexagon is an actual cable line, and the impedance is determined by the line parameters.

[0072] Understandably, this step involves deploying smart power exchange base stations within each energy storage subgrid at the vertices of a regular hexagon, forming hexagonal cellular units. The central node is used to deploy a grid-type energy storage converter as the subgrid control core, while the vertex nodes connect to new energy sources or loads. Each side of the hexagon corresponds to an actual cable line. This deployment method utilizes the symmetrical geometry of the hexagon to naturally offset circulating currents, improve fault tolerance, and provide a regular topological foundation for subsequent admittance matrix construction and coordinated control.

[0073] Step 1012: Construct a main ring network and an auxiliary ring network between each energy storage subnetwork. The main ring network connects the central nodes of each subnetwork, and the auxiliary ring network forms virtual connections between vertex nodes through Kron reduction. Establish the admittance matrix of the main ring network based on the line impedance parameters, and establish the admittance matrix of the auxiliary ring network based on Kron reduction.

[0074] Specifically, a dual-ring network is generated, including a main ring network connecting the SPIEES central node, where the admittance matrix... Energy backbone transport derived from Maxwell equations; admittance matrix of main ring network .

[0075] (1) Physical basis: for any node Kirchhoff's Current Law (KCL):

[0076] ;

[0077] ;

[0078] In the formula, This represents the total current at node i. This indicates that the convergence will occur at node i. Represents the voltage at node i. Represents the voltage at node j. This represents the current between nodes i and j. This represents the mutual admittance between nodes i and j.

[0079] (2) Discretization processing (based on lines) For example, the formula is as follows:

[0080] ;

[0081] In the formula, This represents the line resistance between nodes i and j. This represents the line inductance between nodes i and j.

[0082] (3) The matrix is ​​constructed as follows:

[0083] ;

[0084] In the formula, The diagonal elements are the sum of the admittances of the connected branches.

[0085] Auxiliary ring network admittance matrix The reduced virtual inertia channel admittance matrix is ​​constructed as follows:

[0086] Assume the total number of nodes in the system is (n SPIES centers + m vertex nodes):

[0087] (4) Definition of block matrix:

[0088] ;

[0089] In the formula, This represents the mutual admittance between central nodes (main diagonal) and the admittance connecting to vertex nodes (non-diagonal). This represents the mutual admittance (sparse matrix) from the center node to the vertex node.

[0090] (5) Kron reduction derivation:

[0091] Node voltage equations: ;

[0092] eliminate : ;

[0093] Auxiliary ring network: Provides virtual inertia support, using a Kron-reduced network. The node admittance matrix is ​​simplified as follows: Using the Kron-reduced equations, the virtual inertia channel is constructed as follows:

[0094] ;

[0095]

[0096] An equivalent virtual inertia transfer network is constructed by eliminating internal nodes (m) and retaining boundary nodes (n SPIES centers).

[0097] The main ring network, represented by solid lines, connects all central nodes and is used to transmit active power. The auxiliary ring network, represented by dashed lines, forms virtual connections between vertex nodes through Kronal mass reduction and is used to transmit virtual inertia signals. Please refer to [link to relevant documentation]. Figure 2 .

[0098] Understandably, this step describes the specific method of constructing a dual-ring network between the various energy storage subnetworks: the main ring network connects the central nodes of all subnetworks, establishing an admittance matrix based on Kirchhoff's current law and line impedance parameters to achieve the backbone transmission of active power; the auxiliary ring network eliminates internal nodes (such as renewable energy units and load nodes) through Kron reduction, retaining only the central node as the boundary, and constructs an equivalent virtual inertia transfer network to form virtual connections between vertex nodes and transmit virtual inertia signals. The main ring network is represented by a solid line, and the auxiliary ring network by a dashed line. This structure, through the separation of physical lines and virtual inertia channels, ensures the reliability of power transmission and enhances the system's ability to withstand transient frequency fluctuations.

[0099] Step 1013: Generate a topology connection matrix based on the actual connection relationship between each node. The positions of directly connected nodes are recorded as connected states, and otherwise as non-connected states.

[0100] Specifically, the topology connectivity matrix ,include:

[0101] The data structure is a sparse matrix (COO format).

[0102] The generation rule is: nodes must be directly connected; otherwise:

[0103] ;

[0104] Assemble the admittance matrix:

[0105] ;

[0106] Understandably, this step generates a topology connection matrix based on the actual physical connections between nodes: if two nodes are directly connected, the corresponding position is marked as 1 (connected state); otherwise, it is marked as 0 (unconnected state). This matrix is ​​stored in a sparse matrix format to reduce computational complexity. The purpose of this matrix is ​​to provide a topological basis for the subsequent assembly of the admittance matrix, i.e., the admittance between directly connected nodes is non-zero, which is then used to construct the electrical connection relationships of the system, supporting power flow calculations and dynamic simulations.

[0107] Step 1014: Based on the main ring network admittance matrix, the auxiliary ring network admittance matrix, and the injected current, the initial node voltage is obtained by solving the node voltage equation.

[0108] Specifically, initial voltage / frequency ,include:

[0109] Solve for the initial state (node ​​voltage equations):

[0110] ;

[0111] In the formula, Indicates the nominal voltage.

[0112] Output: Initial node voltage Topological connection matrix , (Fundamentals of dynamic interaction).

[0113] The output generation logic for steps 1013 and 1014 is shown in the table below.

[0114]

[0115] Understandably, this step, based on the established admittance matrices of the main ring network and the auxiliary ring network, as well as the injected current, obtains the initial voltage of each node by simultaneously solving the node voltage equations. The main ring network and the auxiliary ring network jointly participate in the equation solution, with the injected current serving as a known boundary condition. This calculation provides initial values ​​for the system's steady-state operation and forms the basis for subsequent quantum field cooperative control (step 102) calculations of power flow and the superposition of virtual impedance in asymmetric dynamic impedance reshaping (step 103), ensuring that the dynamic simulation starts from a reasonable physical initial state.

[0116] The advantages of this embodiment compared to the traditional topology are shown in the table below.

[0117]

[0118] The specific process of comparing and analyzing this embodiment with traditional topology is as follows:

[0119] I. Calculation Principles and Steps:

[0120] 1. Side length relationship of a single hexagonal unit: circumcircle radius Distance from a vertex of the hexagon to its geometric center; side length : Spacing between adjacent vertices (i.e., the actual length of the cable).

[0121] The properties of a regular hexagon are: ;

[0122] Geometric proof: A regular hexagon can be divided into 6 equilateral triangles, therefore the side length is... .

[0123] 2. Total length of internal connections within the unit: From center to vertex: 6 radial lines, each with a length of... →Total length Vertex connections: 6 edges, each with a length of... →Total length .

[0124] Total length of the unit's internal center vertex:

[0125] ;

[0126] 3. Multi-unit interconnection optimization (key innovation);

[0127] In a honeycomb topology, adjacent cells share the connections between vertices: the length of each connection between vertices (...) Shared by two units → the actual length to be counted needs to be halved.

[0128] The effective vertex length is:

[0129] ;

[0130] Interconnection of central nodes: The main ring network connects the centers of adjacent units, with spacing between them.

[0131] The formula for the distance between the centers is:

[0132] ;

[0133] Derivation: Distance between two centers = Geometric distance between two elements = .

[0134] 4. The total line length of N units is shown in the table below;

[0135]

[0136] Total length formula:

[0137] ;

[0138] Large power grid simplification ( (ignoring boundary effects)

[0139] The average length per unit is:

[0140] ;

[0141] II. Comparison with traditional star topology is as follows:

[0142] 1. Star topology line length: from center to 6 vertices: length of each line → Total length (No connections between vertices)

[0143] 2. The increased length is due to the additional costs incurred by the honeycomb topology, as follows:

[0144] Vertex connections: Add (After sharing);

[0145] Inter-center interconnection: increase ;

[0146] The cost is:

[0147] ;

[0148] 3. The core advantages gained in the exchange are shown in the table below.

[0149]

[0150] In one embodiment, step 102 includes:

[0151] Step 1021: Construct a quantum state Hamiltonian based on the initial node voltage, the initial angular frequency, the main ring network admittance matrix, and the auxiliary ring network admittance matrix. The Hamiltonian includes a kinetic energy term, a voltage potential energy term, and a frequency coupling potential energy term.

[0152] Specifically: Input: , , , (Fundamentals of Dynamic Interaction);

[0153] Constructing the quantum state Hamiltonian as follows:

[0154] ;

[0155]

[0156] The voltage potential energy frequency coupling potential energy is:

[0157] ;

[0158] In the formula, Represents the voltage stiffness coefficient, derived from Diagonal elements determine:

[0159] Represents the coupling coefficient between nodes, proportional to Off-diagonal elements: (Admittance magnitude normalized to s / rad²) Indicates the reference voltage.

[0160] Right now:

[0161] ;

[0162] Understandably, this step constructs a quantum state Hamiltonian containing kinetic energy, voltage potential energy, and frequency coupling potential energy terms based on the initial node voltages, angular frequencies, main loop admittance matrix, and auxiliary loop admittance matrix. The kinetic energy term is defined by the node power deviation, reflecting energy storage inertia; the voltage potential energy term is determined by the main loop admittance and used to constrain voltage deviation; and the frequency coupling potential energy term is determined by the auxiliary loop admittance, characterizing the frequency coupling relationship between nodes. This Hamiltonian dynamically maps the power grid to a quantum system, providing a unified energy operator for subsequent solving of the Schrödinger equation and realizing circulating current suppression and coordinated control based on the quantum tunneling effect.

[0163] Step 1022: Determine the initial quantum state amplitude of each subnet based on the topological connection matrix, and solve the Schrödinger equation using the Hamiltonian to obtain the evolution result of the joint quantum state of the multi-subnet over time; specifically: Solve the Schrödinger equation:

[0164] .

[0165] Understandably, this step determines the initial quantum state amplitude of each subnet based on the topological connectivity matrix (usually determined by the ratio of node voltage to rated voltage), and solves the Schrödinger equation using the constructed Hamiltonian, thus obtaining the evolution of the joint quantum state of the multi-subnet over time. The topological connectivity matrix is ​​used to identify the coupling relationships between adjacent subnets, the initial amplitude reflects the current electrical state of each subnet, and the Schrödinger equation, as the core equation of quantum dynamics, describes the evolution of the system state driven by the Hamiltonian. This evolution result provides subsequent steps with a joint quantum state containing transient information, used for quantum expectation value inversion to generate cooperative control commands, and for calculating the quantum tunneling probability to switch operating modes, thereby achieving circulation suppression and fast response.

[0166] Step 1023: Based on the evolution of the multi-subnet joint quantum state over time, calculate the voltage correction of each node through quantum expectation value inversion, and generate cooperative control commands;

[0167] Specifically, the output coordinated control command is as follows:

[0168] ;

[0169]

[0170] Understandably, this step calculates the voltage correction of each node based on the evolution of the joint quantum state of the multi-subnet over time through quantum expectation value inversion (i.e., measuring the expectation value of the quantum state), thereby generating a cooperative control command. The initial quantum state amplitude is determined by the topological connection matrix and the initial node voltages, reflecting the initial electrical state of the subnet; the evolved joint quantum state contains cooperative information during the system's transient process; expectation value inversion transforms the superposition information of the quantum states into an executable voltage correction. This control command is used in step 103 to adjust the dynamic admittance of each node, achieving rapid cooperative control of the system's transient response.

[0171] Step 1024: Based on the evolution of the joint quantum state of the multi-subnet over time, calculate the quantum tunneling probability of each subnet. When the quantum tunneling probability is less than a preset threshold, switch the working mode of the intelligent power switching base station in that subnet and generate an intelligent power switching base station working mode vector.

[0172] Specifically, quantum tunneling suppression circulation includes: calculating the SPIES mode switching probability ( =Voltage source, =Current source) as follows: (Time switching mode)

[0173] ;

[0174]

[0175]

[0176] Understandably, this step calculates the quantum tunneling probability of each subnet based on the evolution of the joint quantum state over time, where the circulating current energy threshold is modeled as the barrier height. When the tunneling probability falls below a preset threshold, the intelligent power switching base station switches its operating mode from a voltage source to a current source (or vice versa), thereby generating an operating mode vector. Its purpose is to actively suppress circulating current using the quantum tunneling effect, achieving dynamic decoupling and coordination among multiple subnets, and preventing the accumulation of circulating current energy that could lead to system instability.

[0177] Finally, the following commands are output: the collaborative control command from step 1013. and the SPIES working mode vector from step 1014 .

[0178] In one embodiment, step 103 includes:

[0179] Step 1031: Extract the voltage correction amount of each node from the cooperative control command, and calculate the admittance gain coefficient of each node based on the voltage correction amount;

[0180] Specifically, enter: , (From step 102); Real-time frequency change rate (Based on local measurements).

[0181] The processing flow is as follows:

[0182] Firstly, from Extract key parameters from;

[0183] Voltage correction amount → Calculate admittance gain coefficient :

[0184] ;

[0185] Understandably, this step extracts the voltage correction values ​​for each node from the cooperative control commands and calculates the admittance gain coefficient for each node accordingly. The voltage correction values ​​reflect the voltage adjustment requirements given by the quantum field cooperative control, and the admittance gain coefficient converts them into amplitude adjustment factors for dynamic admittance. These factors are used in subsequent steps to dynamically adjust the admittance characteristics of the nodes, thereby achieving adaptive control of the system's transient response and circulating current suppression.

[0186] Step 1032: Determine the asymmetric delay factor and master-slave weight coefficient of each node based on the working mode vector of the intelligent power switching base station;

[0187] Specifically, pattern markers →Determine the direction of delay:

[0188] The voltage source mode and current source mode (operating modes of intelligent power switching base stations) are as follows:

[0189] ;

[0190] Generate the weight matrix, including:

[0191] Construct a diagonal weight matrix :

[0192] The master-slave weighting coefficient is then:

[0193] ;

[0194]

[0195] Understandably, this step determines the asymmetric delay factor of each node based on the operating mode (voltage source or current source) of the intelligent power switching base station, and determines the master-slave weight coefficients based on the master-slave role (master or slave). Different operating modes correspond to different communication delay characteristics; the voltage source mode has a larger delay, while the current source mode has a smaller delay. The master node has a higher weight, and the slave node has a lower weight. These parameters are used for subsequent dynamic admittance calculations and weighted synthesis of the time-varying impedance matrix, achieving dynamic impedance reshaping through master-slave collaboration and improving the system's adaptive adjustment capability to circulating current and transient response.

[0196] Step 1033: Calculate the dynamic admittance of each node based on the admittance gain coefficient, the asymmetric delay factor, and the preset lead time constant and lag time constant.

[0197] Specifically, the single-node admittance (i.e., the dynamic admittance of a node) is calculated as follows:

[0198] ;

[0199] Understandably, this step calculates the dynamic admittance of each node based on the admittance gain coefficient, the asymmetric delay factor, and preset lead and lag time constants. The expression for the dynamic admittance combines a lead-lag element with an asymmetric delay term: the lead time constant improves transient response speed, the lag time constant filters out high-frequency noise, the asymmetric delay factor simulates communication or control delay, and the admittance gain coefficient adaptively adjusts its amplitude according to the voltage correction. This dynamic admittance changes in real time with the system state, providing a basis for subsequent conversion to a time-varying impedance matrix, thereby achieving asymmetric dynamic impedance reshaping, effectively suppressing circulating currents, and improving transient support capabilities.

[0200] Step 1034: Convert the dynamic admittance of each node into dynamic impedance, and then synthesize them by weighting the master-slave weight coefficients of each node to obtain the time-varying impedance matrix.

[0201] Specifically, admittance to impedance transformation (i.e., i.e., variable impedance matrix) )as follows:

[0202] ;

[0203] Frequency domain to time domain discretization is as follows:

[0204] Using the Tustin transform (bilinear approximation), it can be expressed as:

[0205] ;

[0206] Asymmetric delay term Convert to FIR filter:

[0207] other:

[0208] ;

[0209] Full matrix synthesis:

[0210] ;

[0211] In the formula, Indicates the first Unit vector at each node, This represents the inverse Z-transform (real-time DSP implementation).

[0212]

[0213] Understandably, this step first converts the dynamic admittance of each node into dynamic impedance by taking the reciprocal. Then, it discretizes the frequency domain model into a time-domain difference equation through a bilinear transformation, approximating the asymmetric delay term as a finite impulse response filter. Finally, it uses the master-slave weight coefficients of each node to perform a weighted summation, synthesizing the time-varying impedance matrix of the entire system. This matrix can be updated in real time with the system's operating state, providing impedance parameters for subsequent master-slave game optimization. This allows the impact of dynamic impedance on loss and stability to be considered in power scheduling, achieving synergistic optimization of economy and stability.

[0214] In one embodiment, step 104 includes:

[0215] Step 1041: Calculate the power transmission loss coefficient of each node based on the time-varying impedance matrix, and construct an upper-level global optimization model in conjunction with the electricity price signal. With the goal of minimizing the electricity purchase cost and risk cost, solve for the total scheduling power of each node.

[0216] Specifically, first input: time-varying impedance matrix (From step 103); Electricity price signal The subsequent processing procedure is as follows:

[0217] The role of dynamic impedance in two-layer optimization is as follows:

[0218] Input dynamic impedance matrix Directly affects the power transmission loss and stability constraints of the master-slave game model:

[0219] Upper-level global optimization: Calculate the tie-line power loss factor using impedance values. The electricity purchase cost item is revised as follows:

[0220] ;

[0221]

[0222] In the two-layer optimization modeling, the upper layer (global risk control) is:

[0223] ;

[0224]

[0225] Understandably, this step calculates the power transmission loss coefficient of each node based on the time-varying impedance matrix and constructs an upper-level global optimization model by combining real-time electricity price signals. The model aims to minimize the weighted sum of electricity purchase cost and conditional value at risk (CVaR), where the electricity purchase cost is adjusted for power transmission loss using loss coefficients and dynamic impedance magnitudes, and the risk cost is used to quantify expected losses in extreme scenarios. By solving this model, the total dispatch power of each node is obtained, providing global economic and risk control constraints for lower-level optimization, achieving synergistic optimization of stability and economy.

[0226] Step 1042: Calculate the positive-sequence impedance and negative-sequence impedance of each node based on the time-varying impedance matrix, and then determine the charging and discharging efficiency of each energy storage; combine the total scheduling power and the charging and discharging efficiency of each energy storage to construct a lower-level local optimization model, with the goal of maximizing the revenue of each node, and determine the charging and discharging state variables of each energy storage.

[0227] Specifically, lower-level local optimization: positive / negative sequence impedance values ​​after impedance reshaping ( , Impact on energy storage output efficiency (i.e., the charging and discharging efficiency of each energy storage unit), and then adjust the revenue term in the objective function as follows: (where the basic energy storage efficiency is)

[0228] ;

[0229] The lower layer (local revenue optimization) consists of:

[0230] ;

[0231] In the formula, The charging and discharging efficiency of energy storage unit k is represented by (from (Dynamic calculation) This represents the control deviation penalty coefficient (usually taken as 0.01~0.1). This indicates a reference control instruction (from upper-level optimization).

[0232]

[0233] Understandably, the steps involve calculating the positive and negative sequence impedances of each node based on the time-varying impedance matrix, thereby determining the charging and discharging efficiency of each energy storage unit (efficiency is determined by the difference between positive and negative sequence impedances and the basic efficiency), and constructing a lower-level local optimization model by combining the total scheduling power given by the upper-level optimization. The lower-level model aims to maximize the revenue of each node, where the revenue term consists of the product of the charging and discharging efficiency and the scheduling power, while introducing a control deviation penalty term to suppress policy deviations from the reference command. By solving this model, the charging and discharging state variables (such as charging and discharging states) of each energy storage unit are obtained, thus achieving coordinated optimization of local economic benefits and system stability under the influence of dynamic impedance.

[0234] Step 1043: Transform the upper-level global optimization model and the lower-level local optimization model into the Ising model. The objective function of the Ising model includes a quadratic term and a linear term composed of the charging and discharging state variables.

[0235] Specifically, quantum annealing is used to solve for Nash equilibrium as follows:

[0236] Transform the game objective into the Ising model: (representing charging and discharging states)

[0237] Total system energy (objective function value) as follows:

[0238]

[0239] In the formula, Indicates energy storage and discharge; Indicates energy storage charging. This represents the coupling strength between nodes i and j (related to impedance difference and power demand). This represents the local potential field at node i (including electricity price and impedance cost).

[0240] Impedance coupling in quantum annealing solutions includes:

[0241] Embedding the impedance matrix into the coupling coefficients of the Ising model and local field Coupling coefficient: characterizes the energy interaction constraints between nodes.

[0242] ;

[0243] In the formula, Let i be the power requirement of node i;

[0244] Local field: reflects the economic efficiency of the node itself;

[0245] ;

[0246]

[0247] Understandably, this step transforms the upper-level global optimization model and the lower-level local optimization model into the Ising model, whose objective function includes quadratic and linear terms composed of the charging and discharging state variables (taking values ​​of +1 or -1) of each energy storage unit. The coupling coefficient of the quadratic term is calculated based on the dynamic impedance difference and power demand between nodes, characterizing the energy interaction constraints between nodes; the local field of the linear term comprehensively considers real-time electricity prices and dynamic impedance costs, reflecting the economic viability of each node. By solving the ground state (i.e., the state with the lowest total system energy) of this Ising model using quantum annealing, the Nash equilibrium solution of the two-layer game can be obtained. The significance of this transformation lies in mapping the economic scheduling and stability constraint problem to the ground state search of a quantum physical system. Utilizing the efficient parallel computing power of quantum annealing, it solves the problem of long solution times in traditional optimization methods, achieving millisecond-level economic scheduling decisions.

[0248] Step 1044: Solve the ground state of the Ising model by quantum annealing to obtain the charging and discharging state sequence of each energy storage, calculate the actual charging and discharging power of each energy storage based on the charging and discharging state sequence, and generate an energy storage power scheduling plan.

[0249] Specifically, the generation of energy storage power dispatch plans includes:

[0250] Upper-level optimization output: Global optimal power allocation To meet the requirements, the cost of electricity purchase is:

[0251]

[0252] Lower-level optimized output: The ground state of the Ising model is solved by quantum annealing to obtain the charge / discharge state sequence. The actual output of each node is then calculated as follows:

[0253] ;

[0254]

[0255] Understandably, this step uses quantum annealing to solve for the ground state of the Ising model (i.e., the stable state with the lowest total system energy), obtaining the charging and discharging state sequence of each energy storage unit (+1 represents discharging, -1 represents charging). Then, based on the rated charging and discharging power of each energy storage unit, the actual charging and discharging power is calculated, ultimately generating an energy storage power scheduling plan. The upper-level optimization provides a globally optimal power allocation as a boundary reference, while the lower-level optimization utilizes the efficient parallel search capability of quantum annealing to quickly find the charging and discharging state that minimizes the total system energy from an exponential number of state combinations. The purpose of this step is to transform the complex two-level game optimization problem into a quantum physics ground state search, significantly reducing the solution time and achieving millisecond- to second-level economic scheduling decisions, meeting the requirements of speed and optimality in high-proportion new energy scenarios.

[0256] Step 1045: Based on the energy storage power scheduling plan, calculate the electricity purchase cost, risk cost, and control deviation penalty, and sum them to obtain the game equilibrium cost.

[0257] Specifically, the equilibrium cost calculation is as follows:

[0258] Nash equilibrium verification:

[0259] Check if the following conditions are met:

[0260] The convergence threshold is usually set to [value missing]. .

[0261] Economic indicators: The risk cost reduction rate is calculated by comparing the CVaR values ​​before and after the game.

[0262] ;

[0263]

[0264] ;

[0265] Game equilibrium cost The calculation formulas include:

[0266] Electricity purchase cost equilibrium risk cost control deviation penalty:

[0267]

[0268]

[0269] Final output: Upper-level global energy storage power scheduling plan Game equilibrium cost .

[0270] Understandably, this step calculates the electricity purchase cost (based on real-time electricity prices and the purchased power considering dynamic impedance losses), risk cost (quantifying expected losses in extreme scenarios using Conditional Value at Risk (CVaR), and control deviation penalty (suppressing the deviation between local control commands and upper-level reference commands) based on the generated energy storage power dispatch plan. The weighted sum of these three costs yields the game equilibrium cost. Simultaneously, the optimization effect is evaluated through Nash equilibrium verification (checking whether the objective function gradient converges) and the risk cost reduction rate. This equilibrium cost, as the final economic indicator, comprehensively reflects the system's electricity purchase expenditure, operational risk, and control accuracy under the influence of dynamic impedance, providing a quantitative basis for the convergence and economic benefits of master-slave game optimization.

[0271] Furthermore, the present invention also provides a comparative verification of the algorithm's advantages as follows:

[0272]

[0273] This algorithm, through a quantum-game collaborative framework, deeply couples physical mechanisms (steps 101-103) with economic dispatch (step 104), providing a full-scale solution for grid-based energy storage with millisecond-level response and minute-level optimization. RT-LAB hardware-in-the-loop testing shows that in scenarios with a renewable energy penetration rate >50%, the system instability risk is reduced by 67%, and electricity purchase costs are lowered.

[0274] In summary, the modeling and economic dispatch method for grid-based energy storage provided by this embodiment of the invention firstly involves deploying intelligent power exchange base stations within each energy storage sub-network according to the coordinates of the vertices of a regular hexagon to form hexagonal units. A dual-ring topology consisting of a main ring network and an auxiliary ring network is then constructed between each sub-network, generating a topology connection matrix, initial node voltages and angular frequencies, and admittance matrices for the main and auxiliary ring networks. The hexagonal symmetric structure naturally cancels circulating currents and improves fault tolerance. The main ring network transmits active power, while the auxiliary ring network constructs virtual inertia channels through Kron reduction, laying a regular electrical foundation for subsequent rapid coordinated control. Then, based on the aforementioned topology and admittance matrix, a quantum state Hamiltonian containing kinetic energy, voltage potential energy, and frequency coupling potential energy terms is constructed. The Schrödinger equation is solved to obtain the multi-sub-network joint quantum state evolution results, thereby executing quantum field coordinated control and generating coordinated control commands and intelligent power exchange base station operating mode vectors. By actively suppressing circulating current using the quantum tunneling effect and rapidly generating voltage correction values ​​through quantum expectation value inversion, the inertia response time is significantly shortened, voltage recovery rate and fault ride-through success rate are improved, and transient response capability is greatly enhanced. Secondly, based on the cooperative control commands and operating mode vectors, the admittance gain coefficient, asymmetric delay factor, and master-slave weight coefficient of each node are calculated, thereby obtaining dynamic admittance and converting it into a time-varying impedance matrix. This process achieves asymmetric dynamic impedance reshaping, enabling impedance parameters to adaptively adjust in real time according to system state, further suppressing circulating current and enhancing dynamic support, effectively improving system robustness. Finally, based on the time-varying impedance matrix and electricity price signal, an upper-level global optimization model (minimizing electricity purchase cost and conditional risk value) and a lower-level local optimization model (maximizing the revenue of each node) are constructed. The two-layer game is transformed into an Ising model, and the Nash equilibrium is quickly solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost. Quantum annealing optimization significantly shortens the solution time, achieving synergy between millisecond-level control and minute-level optimization, effectively reducing electricity purchase cost and operational risk. Therefore, this invention solves the problems of insufficient transient response capability and low optimization scheduling efficiency in the prior art.

[0275] The above describes a modeling and economic dispatch method for grid-connected energy storage provided in an embodiment of the present invention. The following describes a modeling and economic dispatch system for grid-connected energy storage provided in an embodiment of the present invention.

[0276] Please see Figure 3 The present invention provides a modeling and economic dispatch system for grid-connected energy storage, comprising:

[0277] The first computing unit 201 is used to deploy smart power switching base stations according to the hexagonal vertex coordinates within each energy storage subnet to form a hexagonal unit; to construct a double ring topology between each subnet to connect each subnet; and to generate a topology connection matrix, initial node voltage and initial angular frequency, main ring network admittance matrix and auxiliary ring network admittance matrix according to the hexagonal unit and the double ring topology.

[0278] The second computing unit 202 is used to perform quantum field cooperative control based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, by constructing quantum state Hamiltonians and solving the Schrödinger equation, so as to generate cooperative control commands and intelligent power exchange base station operating mode vectors.

[0279] The third calculation unit 203 is used to perform asymmetric dynamic impedance reshaping based on the cooperative control command and the working mode vector of the intelligent power exchange base station, calculate the dynamic admittance of each node, and convert the dynamic admittance into a time-varying impedance matrix.

[0280] The fourth calculation unit 204 is used to perform master-slave game optimization based on the time-varying impedance matrix and combined with the electricity price signal, solve the Nash equilibrium through quantum annealing, and generate an energy storage power scheduling plan and game equilibrium cost.

[0281] In one embodiment, the first computing unit 201 is specifically used for:

[0282] Within each energy storage subnetwork, each smart power exchange base station is deployed according to the vertex coordinates of a regular hexagon, so that each base station constitutes a hexagonal unit.

[0283] Between each energy storage subnetwork, a main ring network and an auxiliary ring network are constructed. The main ring network connects the central nodes of each subnetwork, and the auxiliary ring network forms virtual connections between vertex nodes through Kron reduction. The admittance matrix of the main ring network is established based on the line impedance parameters, and the admittance matrix of the auxiliary ring network is established based on Kron reduction.

[0284] Based on the actual connection relationships between each node, a topology connection matrix is ​​generated, where the positions of directly connected nodes are marked as connected states, and those not connected are marked as non-connected states.

[0285] Based on the admittance matrix of the main ring network, the admittance matrix of the auxiliary ring network, and the injected current, the initial node voltage is obtained by solving the node voltage equation.

[0286] In one embodiment, the second computing unit 202 is specifically used for:

[0287] Based on the initial node voltage, the initial angular frequency, the main ring network admittance matrix, and the auxiliary ring network admittance matrix, a quantum state Hamiltonian is constructed, which includes a kinetic energy term, a voltage potential energy term, and a frequency coupling potential energy term.

[0288] The initial quantum state amplitude of each subnet is determined based on the topological connection matrix, and the Schrödinger equation is solved by combining the Hamiltonian to obtain the evolution result of the joint quantum state of multiple subnets over time.

[0289] Based on the evolution of the multi-subnet joint quantum state over time, the voltage correction of each node is calculated by quantum expectation value inversion, and cooperative control commands are generated.

[0290] Based on the evolution of the joint quantum state of the multi-subnet over time, the quantum tunneling probability of each subnet is calculated. When the quantum tunneling probability is less than a preset threshold, the working mode of the intelligent power switching base station in that subnet is switched, and an intelligent power switching base station working mode vector is generated.

[0291] In one embodiment, the third computing unit 203 is specifically used for:

[0292] The voltage correction amount of each node is extracted from the cooperative control command, and the admittance gain coefficient of each node is calculated based on the voltage correction amount;

[0293] Based on the working mode vector of the intelligent power switching base station, determine the asymmetric delay factor and master-slave weight coefficient of each node;

[0294] The dynamic admittance of each node is calculated based on the admittance gain coefficient, the asymmetric delay factor, and the preset lead time constant and lag time constant.

[0295] The dynamic admittance of each node is converted into dynamic impedance, and then weighted and synthesized using the master-slave weight coefficients of each node to obtain the time-varying impedance matrix.

[0296] In one embodiment, the fourth calculation unit 204 is specifically used for:

[0297] The power transmission loss coefficient of each node is calculated based on the time-varying impedance matrix, and an upper-level global optimization model is constructed in conjunction with the electricity price signal. With the goal of minimizing the electricity purchase cost and risk cost, the total scheduling power of each node is obtained.

[0298] The positive-sequence impedance and negative-sequence impedance of each node are calculated based on the time-varying impedance matrix, thereby determining the charging and discharging efficiency of each energy storage. Combined with the total scheduling power and the charging and discharging efficiency of each energy storage, a lower-level local optimization model is constructed to determine the charging and discharging state variables of each energy storage with the goal of maximizing the revenue of each node.

[0299] The upper-level global optimization model and the lower-level local optimization model are transformed into the Ising model. The objective function of the Ising model includes a quadratic term and a linear term composed of the charging and discharging state variables.

[0300] The ground state of the Ising model is solved by quantum annealing to obtain the charging and discharging state sequence of each energy storage unit. The actual charging and discharging power of each energy storage unit is calculated based on the charging and discharging state sequence to generate an energy storage power scheduling plan.

[0301] Based on the energy storage power dispatch plan, the electricity purchase cost, risk cost, and control deviation penalty are calculated and summed to obtain the game equilibrium cost.

[0302] This invention provides a modeling and economic dispatch system for grid-based energy storage. First, within each energy storage subgrid, intelligent power exchange base stations are deployed according to the coordinates of the vertices of a regular hexagon to form hexagonal units. A dual-ring topology consisting of a main ring network and an auxiliary ring network is constructed between the subgrids. Based on this, a topology connection matrix, initial node voltages and angular frequencies, and admittance matrices for the main and auxiliary ring networks are generated. The hexagonal symmetric structure naturally cancels circulating currents and improves fault tolerance. The main ring network transmits active power, and the auxiliary ring network constructs virtual inertia channels through Kron reduction, laying a regular electrical foundation for subsequent rapid coordinated control. Then, based on the above topology and admittance matrix, a quantum state Hamiltonian containing kinetic energy, voltage potential energy, and frequency coupling potential energy terms is constructed. The Schrödinger equation is solved to obtain the multi-subgrid joint quantum state evolution results, and then quantum field coordinated control is executed to generate coordinated control commands and intelligent power exchange base station operating mode vectors. By actively suppressing circulating current using the quantum tunneling effect and rapidly generating voltage correction values ​​through quantum expectation value inversion, the inertia response time is significantly shortened, voltage recovery rate and fault ride-through success rate are improved, and transient response capability is greatly enhanced. Secondly, based on the cooperative control commands and operating mode vectors, the admittance gain coefficient, asymmetric delay factor, and master-slave weight coefficient of each node are calculated, thereby obtaining dynamic admittance and converting it into a time-varying impedance matrix. This process achieves asymmetric dynamic impedance reshaping, enabling impedance parameters to adaptively adjust in real time according to system state, further suppressing circulating current and enhancing dynamic support, effectively improving system robustness. Finally, based on the time-varying impedance matrix and electricity price signal, an upper-level global optimization model (minimizing electricity purchase cost and conditional risk value) and a lower-level local optimization model (maximizing the revenue of each node) are constructed. The two-layer game is transformed into an Ising model, and the Nash equilibrium is quickly solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost. Quantum annealing optimization significantly shortens the solution time, achieving synergy between millisecond-level control and minute-level optimization, effectively reducing electricity purchase cost and operational risk. In summary, this invention solves the problems of insufficient transient response capability and low optimization scheduling efficiency in the prior art.

[0303] Furthermore, this embodiment of the invention also provides a modeling and economic dispatching device for grid-connected energy storage, the device comprising a processor and a memory:

[0304] The memory is used to store program code and transmit the program code to the processor;

[0305] The processor is used to execute the steps of the modeling and economic dispatch method for grid-based energy storage as described in the above method embodiments, according to the instructions in the program code.

[0306] Furthermore, this embodiment of the invention also provides a computer-readable storage medium for storing program code, which is used to execute the modeling and economic dispatch method for grid-based energy storage described in the above method embodiments.

[0307] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0308] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0309] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0310] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0311] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0312] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A modeling and economic dispatch method for grid-connected energy storage, characterized in that, include: Within each energy storage subnetwork, intelligent power switching base stations are deployed according to the coordinates of the hexagonal vertices to form hexagonal units; A dual-ring topology is constructed between the subnets to connect them; based on the hexagonal cells and the dual-ring topology, a topology connection matrix, initial node voltages and initial angular frequencies, a main ring network admittance matrix, and an auxiliary ring network admittance matrix are generated. Based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, quantum field cooperative control is performed by constructing quantum state Hamiltonians and solving the Schrödinger equation to generate cooperative control commands and intelligent power exchange base station operating mode vectors. Based on the cooperative control command and the working mode vector of the intelligent power switching base station, asymmetric dynamic impedance reshaping is performed, the dynamic admittance of each node is calculated, and the dynamic admittance is converted into a time-varying impedance matrix. Based on the time-varying impedance matrix and combined with the electricity price signal, master-slave game optimization is performed, and the Nash equilibrium is solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost.

2. The modeling and economic dispatch method for grid-connected energy storage according to claim 1, characterized in that, Within each energy storage subnetwork, intelligent power switching base stations are deployed according to the coordinates of hexagonal vertices to form hexagonal units; a double ring topology is constructed between each subnetwork to connect them. Based on the hexagonal cells and the double-ring topology, a topology connection matrix, initial node voltages and initial angular frequencies, a main ring network admittance matrix, and an auxiliary ring network admittance matrix are generated, including: Within each energy storage subnetwork, each smart power exchange base station is deployed according to the vertex coordinates of a regular hexagon, so that each base station constitutes a hexagonal unit. Between each energy storage subnetwork, a main ring network and an auxiliary ring network are constructed. The main ring network connects the central nodes of each subnetwork, and the auxiliary ring network forms virtual connections between vertex nodes through Kron reduction. The admittance matrix of the main ring network is established based on the line impedance parameters, and the admittance matrix of the auxiliary ring network is established based on Kron reduction. Based on the actual connection relationships between each node, a topology connection matrix is ​​generated, where the positions of directly connected nodes are marked as connected states, and those not connected are marked as non-connected states. Based on the admittance matrix of the main ring network, the admittance matrix of the auxiliary ring network, and the injected current, the initial node voltage is obtained by solving the node voltage equation.

3. The modeling and economic dispatch method for grid-connected energy storage according to claim 1, characterized in that, The quantum field cooperative control, based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, is executed by constructing a quantum state Hamiltonian and solving the Schrödinger equation to generate cooperative control commands and a smart power exchange base station operating mode vector, including: Based on the initial node voltage, the initial angular frequency, the main ring network admittance matrix, and the auxiliary ring network admittance matrix, a quantum state Hamiltonian is constructed, which includes a kinetic energy term, a voltage potential energy term, and a frequency coupling potential energy term. The initial quantum state amplitude of each subnet is determined based on the topological connection matrix, and the Schrödinger equation is solved by combining the Hamiltonian to obtain the evolution result of the joint quantum state of multiple subnets over time. Based on the evolution of the multi-subnet joint quantum state over time, the voltage correction of each node is calculated by quantum expectation value inversion, and cooperative control commands are generated. Based on the evolution of the joint quantum state of the multi-subnet over time, the quantum tunneling probability of each subnet is calculated. When the quantum tunneling probability is less than a preset threshold, the working mode of the intelligent power switching base station in that subnet is switched, and an intelligent power switching base station working mode vector is generated.

4. The modeling and economic dispatch method for grid-connected energy storage according to claim 3, characterized in that, The expression for the Hamiltonian of the quantum state is: ； In the formula, The Hamiltonian of the quantum state is... For nodes The generalized momentum operator, For node equivalent quality, Let be the potential energy function.

5. The modeling and economic dispatch method for grid-connected energy storage according to claim 1, characterized in that, The process of performing asymmetric dynamic impedance reshaping based on the cooperative control commands and the intelligent power switching base station operating mode vector, calculating the dynamic admittance of each node, and converting the dynamic admittance into a time-varying impedance matrix includes: The voltage correction amount of each node is extracted from the cooperative control command, and the admittance gain coefficient of each node is calculated based on the voltage correction amount; Based on the working mode vector of the intelligent power switching base station, determine the asymmetric delay factor and master-slave weight coefficient of each node; The dynamic admittance of each node is calculated based on the admittance gain coefficient, the asymmetric delay factor, and the preset lead time constant and lag time constant. The dynamic admittance of each node is converted into dynamic impedance, and then weighted and synthesized using the master-slave weight coefficients of each node to obtain the time-varying impedance matrix.

6. The modeling and economic dispatch method for grid-connected energy storage according to claim 5, characterized in that, The expression for calculating the dynamic admittance is: ； In the formula, Let k be the dynamic admittance of node k. The admittance gain coefficient, The lead time constant, The time lag constant is It is an asymmetric delay factor. It is a Laplace complex variable.

7. The modeling and economic dispatch method for grid-connected energy storage according to claim 1, characterized in that, Based on the time-varying impedance matrix and combined with the electricity price signal, master-slave game optimization is performed. The Nash equilibrium is solved through quantum annealing to generate an energy storage power scheduling plan and game equilibrium cost, including: The power transmission loss coefficient of each node is calculated based on the time-varying impedance matrix, and an upper-level global optimization model is constructed in conjunction with the electricity price signal. With the goal of minimizing the electricity purchase cost and risk cost, the total scheduling power of each node is obtained. The positive-sequence impedance and negative-sequence impedance of each node are calculated based on the time-varying impedance matrix, thereby determining the charging and discharging efficiency of each energy storage. Combined with the total scheduling power and the charging and discharging efficiency of each energy storage, a lower-level local optimization model is constructed to determine the charging and discharging state variables of each energy storage with the goal of maximizing the revenue of each node. The upper-level global optimization model and the lower-level local optimization model are transformed into the Ising model. The objective function of the Ising model includes a quadratic term and a linear term composed of the charging and discharging state variables. The ground state of the Ising model is solved by quantum annealing to obtain the charging and discharging state sequence of each energy storage unit. The actual charging and discharging power of each energy storage unit is calculated based on the charging and discharging state sequence to generate an energy storage power scheduling plan. Based on the energy storage power dispatch plan, the electricity purchase cost, risk cost, and control deviation penalty are calculated and summed to obtain the game equilibrium cost.

8. A modeling and economic dispatch system for grid-connected energy storage, characterized in that, include: The first computing unit is used to deploy smart power exchange base stations according to the coordinates of the vertices of a hexagon to form a hexagonal unit within each energy storage subnetwork; A dual-ring topology is constructed between the subnets to connect them; based on the hexagonal cells and the dual-ring topology, a topology connection matrix, initial node voltages and initial angular frequencies, a main ring network admittance matrix, and an auxiliary ring network admittance matrix are generated. The second computing unit is used to perform quantum field cooperative control by constructing quantum state Hamiltonians and solving the Schrödinger equation based on the topology connection matrix, the initial node voltage and the initial angular frequency, the main ring network admittance matrix and the auxiliary ring network admittance matrix, so as to generate cooperative control commands and intelligent power exchange base station operating mode vectors. The third calculation unit is used to perform asymmetric dynamic impedance reshaping based on the cooperative control command and the working mode vector of the intelligent power exchange base station, calculate the dynamic admittance of each node, and convert the dynamic admittance into a time-varying impedance matrix. The fourth calculation unit is used to perform master-slave game optimization based on the time-varying impedance matrix and the electricity price signal, solve the Nash equilibrium through quantum annealing, and generate an energy storage power scheduling plan and game equilibrium cost.

9. A modeling and economic dispatching device for grid-connected energy storage, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the modeling and economic dispatch method for grid-connected energy storage as described in any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the modeling and economic dispatch method for grid-based energy storage as described in any one of claims 1-7.

Images