Power transmission line wide-area optimization scheduling method and device based on quantum computing and digital twinning, and storage medium

Abstract

This invention relates to a method, apparatus, and storage medium for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins, belonging to the field of power grid dispatching technology. Technical solution: Construct a high-fidelity digital twin model of the physical entity of the transmission line, synchronizing the physical power grid status in real time; simultaneously, model the complex wide-area optimal scheduling problem of the power grid into an optimization model suitable for solving by quantum computers; utilize the powerful parallel computing capabilities of quantum computing to rapidly solve and simulate the scheduling scheme in the digital twin environment; finally, issue the optimized scheduling instructions to the physical power grid for execution, achieving globally optimal and dynamically adaptive intelligent scheduling. Positive effects of this invention: Significantly improved computational speed, enhanced decision-making quality (global optimization), improved power grid resilience and intelligence level, and solves the problems of low computational efficiency, poor real-time performance, and difficulty in handling uncertainties in large-scale, complex power grid environment optimization scheduling of transmission lines in existing technologies.

Description

Technical Field

[0001] This invention relates to a method, apparatus, and storage medium for wide-area optimal scheduling of power transmission lines based on quantum computing and digital twins, belonging to the field of power grid scheduling technology. Background Technology

[0002] The power system is an indispensable infrastructure of modern society, and one of its core tasks is to ensure the safe, stable, and economical transmission of electricity. With the large-scale grid connection of new energy sources (wind power and photovoltaics) and the expansion of grid interconnection, the uncertainty, volatility, and complexity of grid operation have increased dramatically. Traditional grid dispatch relies on energy management systems (EMS) based on classical computers and advanced application software (such as optimal power flow calculation). These methods often face challenges such as excessive computation time, difficulty in finding the global optimum, and insufficient adaptability to uncertainty when dealing with ultra-large-scale, nonlinear, and high-dimensional real-time optimization problems. For example, in a power grid containing tens of thousands of nodes and lines, N-1 security-constrained economic dispatch may take tens of minutes or even hours using classical algorithms, failing to meet the real-time dispatch requirements at the minute or even second level.

[0003] Digital twin technology provides power grids with the ability to monitor the entire grid, simulate, and dynamically predict future scenarios; however, the speed of solving complex optimization problems—the core of decision-making—remains a bottleneck. The rise of quantum computing offers revolutionary hope for solving such computationally intensive problems.

[0004] The closest existing technical solution to this invention is a power grid dispatching system based on classical optimization algorithms and digital simulation. This system typically includes:

[0005] Data Acquisition and Monitoring System (SCADA / PMU): Acquires real-time power grid operating data (voltage, current, power, etc.). Digital Simulation or Preliminary Digital Twin Model: Establishes a physical model of the power grid for state estimation and offline simulation analysis. Classical Optimization Solver: Runs optimization programs based on linear programming (LP), nonlinear programming (NLP), mixed integer programming (MIP), or heuristic algorithms (such as genetic algorithms) on the central dispatch server to solve problems related to economic dispatch and optimal power flow. Dispatch Command Generation and Issuance: Issues and executes commands such as generator output plans and switch states obtained from the solutions. The operational flow of this scheme is as follows: real-time data drives the simulation model update; the optimization solver performs periodic (e.g., every 5-15 minutes) or event-triggered calculations based on the current model state and constraints, and outputs the dispatch scheme.

[0006] The disadvantages of the above-mentioned prior art are:

[0007] The contradiction between computational efficiency and scale: The computational complexity of existing classical optimization algorithms (such as the interior-point method for solving nonlinear optimal power flow) typically increases polynomially or even exponentially with the number of power grid nodes. When the power grid scale expands to the provincial or inter-regional level, the computation time may far exceed the requirements of the scheduling cycle, leading to a lag in scheduling decisions. Difficulty in guaranteeing global optimum: For non-convex, high-dimensional, discrete combinatorial optimization problems (such as unit combination problems involving start-up and shutdown states), classical heuristic algorithms are prone to getting trapped in local optima, while exact algorithms are infeasible due to excessive computational cost. This results in the failure to fully realize the economic and security potential of the scheduling scheme. Insufficient flexibility in handling uncertainty: The output of new energy sources has strong randomness. Existing methods often use scenario-based methods or robust optimization, but this greatly increases the problem scale and computational burden. In rapidly fluctuating environments, it is difficult to generate optimal adaptive scheduling strategies that consider multiple uncertainties in a timely manner. Disconnect between digital twins and optimization decision-making: Existing digital twin or simulation systems are mainly used for "seeing" and "simulating," and their coupling with the optimization decision engine is not tight or efficient enough. Simulation results require manual or simple rule intervention to be fed back to the optimization model, making it difficult to form a rapid closed loop of "perception-optimization-decision-execution". Summary of the Invention

[0008] This invention proposes a method, device, and storage medium for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins. It greatly improves the computing speed, enhances the decision quality (global optimization), improves the resilience and intelligence level of the power grid, and solves the problems of low computing efficiency, poor real-time performance, and difficulty in handling uncertainties in the large-scale and complex power grid environment of existing technologies for optimal scheduling of transmission lines.

[0009] The technical solution of this invention is:

[0010] A method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins is proposed. This method constructs a high-fidelity digital twin model of the physical entities of the transmission lines, synchronizing the physical power grid status in real time. Simultaneously, the complex wide-area optimal scheduling problem of the power grid is modeled as an optimization model suitable for solving by quantum computers. Utilizing the powerful parallel computing capabilities of quantum computing, the scheduling scheme is rapidly solved and simulated in the digital twin environment. Finally, the optimized scheduling instructions are issued to the physical power grid for execution, achieving globally optimal and dynamically adaptive intelligent scheduling.

[0011] The complex power grid wide-area optimization scheduling problem includes power flow optimization, economic scheduling, and security verification; the optimization model is a quadratic unconstrained binary optimization-QUBO model, and the quantum computing is a quantum annealing or gate circuit model.

[0012] It includes the following steps:

[0013] Step 1: Real-time sensing and data upload of the physical power grid

[0014] Smart sensors distributed throughout the power grid collect electrical quantities, equipment status, environmental data, and power prediction data of new energy power plants in real time, and upload the real-time data to the digital twin layer through the communication network.

[0015] Step Two: Construction and Synchronization of Wide-Area Digital Twins for Transmission Lines

[0016] ① Model Construction: In the digital twin layer, based on the power grid GIS, equipment parameters, and physical laws, a high-fidelity, multi-physics coupled digital mirror model is established, covering all transmission lines, substations, generator sets, and loads of the target wide-area power grid. This model can not only reflect steady-state electrical characteristics but also simulate electromagnetic transients, thermal dynamics, and other processes.

[0017] ② Real-time synchronization: Using the real-time data uploaded in step one, the operating status of the digital twin model is dynamically calibrated through technologies such as state estimation and parameter identification, so that it is synchronized with the physical power grid at the millisecond level and truly reflects the current operating condition of the power grid;

[0018] Step 3: Quantum Modeling and Transformation of the Scheduling Problem

[0019] In the quantum-classical hybrid computing layer, scheduling objectives and constraints are received from the application service layer; through a transformation algorithm, the above-mentioned power grid wide-area optimization scheduling problem is mapped or approximately equivalent to a quadratic unconstrained binary optimization QUBO model or the Ising model.

[0020] Step 4: Quantum-Classical Hybrid Solution

[0021] The constructed QUBO model is input into a quantum computing device (such as a quantum annealing machine or a noisy medium-scale quantum-NISQ computer).

[0022] Quantum computing devices utilize quantum parallelism for rapid sampling and search for the lowest energy state, outputting one or more sets of optimized solutions (corresponding to the values ​​of qubits) in a short time.

[0023] Step 5: Scheme Demonstration and Decision Making

[0024] The optimized scheduling scheme and anticipated operation set obtained in step four are injected into the real-time synchronized digital twin model in step two for advanced simulation and N-k safety verification.

[0025] The digital twin quickly calculates the state changes of the power grid in the next scheduling cycle; if the calculation results meet all security constraints, the scheme is confirmed and an optimized scheduling instruction is formed; if a limit violation is found, the new constraints or penalty terms are fed back to step three for iterative optimization until a safe and feasible optimal scheme is obtained and an optimized scheduling instruction is formed.

[0026] Step Six: Issuance and Execution of Commands

[0027] The finalized optimized scheduling instructions are sent to the execution mechanism at the physical power grid layer through a secure channel to complete closed-loop control.

[0028] Furthermore, the intelligent sensor in step one includes a PMU and an intelligent terminal; the electrical quantities of the power transmission line collected include I, V, P, and Q; the equipment status includes switches and transformer taps; and the environmental data includes temperature and wind speed.

[0029] Furthermore, the application service layer in step three receives scheduling objectives including minimizing network loss, minimizing generation cost, and maximizing safety margin, while the constraints include line power flow limits, voltage upper and lower limits, and generator output range; the power grid wide-area optimization scheduling problem typically involves a nonlinear mixed-integer programming problem with discrete variables.

[0030] Furthermore, the transformation algorithm in step three discretizes and encodes continuous variables (such as generator active power output) into a combination of multiple qubits, and encodes the objective function and constraints (including equality constraints and inequality constraints) into linear and quadratic terms in the Hamiltonian of the QUBO model.

[0031] Furthermore, the optimized scheduling instructions in step six include AGC setpoints and switch operation commands, and the actuators include power plant control systems and substation automation systems.

[0032] In this invention, the classical computing coprocessor is responsible for: 1) decoding the original solution output by quantum computing and converting it back into power grid scheduling variables (such as unit start-up and shutdown status, output value); 2) performing feasibility verification and local fine optimization (for example, using the classical QP algorithm to quickly solve the continuous power flow optimization based on the excellent discrete variable combination provided by the quantum solution), so as to make up for the lack of precision and scale of current quantum hardware.

[0033] A wide-area optimization scheduling device for transmission lines based on quantum computing and digital twins comprises a data acquisition and communication module, a digital twin modeling and synchronization module, a problem modeling and transformation module, a quantum-classical hybrid computing module, a scheme verification and decision-making module, and an instruction generation and issuance module. As a hybrid enhanced intelligent scheduling system, it realizes real-time data interaction between layers through a high-speed communication network, and is used to implement the aforementioned wide-area optimization scheduling method for transmission lines based on quantum computing and digital twins.

[0034] Specifically, it includes:

[0035] Data Acquisition and Communication Module: Configured for acquiring real-time data from the physical power grid; Digital Twin Modeling and Synchronization Module: Configured for building and updating the power grid digital twin model; Problem Modeling and Transformation Module: Configured for converting scheduling requirements into quantum optimization models; Quantum-Classical Hybrid Computing Module: Connected to quantum computing devices, responsible for allocating scheduling tasks, receiving and decoding quantum computing results, and classical post-processing; Scheme Verification and Decision Module: Configured for verifying optimization schemes in the digital twin environment; Instruction Generation and Issuance Module: Configured for generating executable optimized scheduling instructions.

[0036] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements all the steps of the aforementioned wide-area optimal scheduling of transmission lines based on quantum computing and digital twins.

[0037] Key points of the invention concept:

[0038] Cross-disciplinary integration of quantum computing and power grid optimization scheduling: The core lies in adapting the classical power system optimization scheduling problem to a quantum computing hardware platform through specific mathematical transformation.

[0039] Digital twins serve as a dynamic verification and iteration platform for quantum optimization results: Digital twins are not only state monitors, but also "sandboxes" for optimization schemes. Their feedback loop with the quantum solver is the key to ensuring the safety and feasibility of the scheme.

[0040] A quantum-classical hybrid collaborative computing architecture: Pure quantum solution is not yet mature at present. This invention designs a hybrid mode in which classical computing (handling continuous variables and accurate verification) and quantum computing (handling core discrete combinatorial optimization) work together, which is practically feasible.

[0041] A real-time scheduling modeling method for wide-area, multi-objective, and strongly constrained scenarios: Design a QUBO modeling method suitable for quantum computing that can comprehensively express multiple scheduling objectives such as economy, security, and environmental protection.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] Direct technical benefits – significantly increased computational speed: Quantum computing has the potential to exponentially speed up certain combinatorial optimization problems. Transforming the complex optimization problem at the core of power grid dispatching into a QUBO model and solving it with a quantum computer could potentially reduce large-scale dispatching calculations that originally took hours to minutes or even seconds.

[0044] Indirect technical effects – enhanced real-time and adaptive scheduling: The increased computing speed enables the scheduling system to perform rolling optimization based on real-time data at shorter time intervals, quickly respond to uncertain events such as fluctuations in renewable energy power and sudden load changes, and achieve near real-time, adaptive wide-area optimized scheduling.

[0045] Indirect technical effects – improved decision quality (global optimization): Algorithms such as quantum annealing tend to search for the global optimal solution or a high-quality approximate solution to a problem. Compared with classical heuristic algorithms that are prone to getting trapped in local optima, they can obtain scheduling schemes with better economic efficiency and higher safety margins, thereby improving the overall operating efficiency of the power grid.

[0046] Ultimate Technical Results – Enhanced Grid Resilience and Intelligence: The deep integration of digital twins and quantum computing forms an intelligent closed loop of "real-time perception – rapid prediction – global optimization – simulation verification – precise execution." This enables the power grid to proactively identify risks and optimize operating strategies, significantly enhancing its resilience in the face of complex faults and uncertainties, representing the development direction of the next generation of smart grid dispatching systems. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0048] A method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins is proposed. This method constructs a high-fidelity digital twin model of the physical entities of the transmission lines, synchronizing the physical power grid status in real time. Simultaneously, the complex wide-area optimal scheduling problem of the power grid is modeled as an optimization model suitable for solving by quantum computers. Utilizing the powerful parallel computing capabilities of quantum computing, the scheduling scheme is rapidly solved and simulated in the digital twin environment. Finally, the optimized scheduling instructions are issued to the physical power grid for execution, achieving globally optimal and dynamically adaptive intelligent scheduling.

[0049] The complex power grid wide-area optimization scheduling problem includes power flow optimization, economic scheduling, and security verification; the optimization model is a quadratic unconstrained binary optimization-QUBO model, and the quantum computing is a quantum annealing or gate circuit model.

[0050] It includes the following steps:

[0051] Step 1: Real-time sensing and data upload of the physical power grid

[0052] Smart sensors (PMUs, smart terminals) deployed throughout the power grid collect real-time electrical quantities (I, V, P, Q) of transmission lines, equipment status (switches, transformer taps), environmental data (temperature, wind speed), and power prediction data of new energy power plants. The real-time data is then uploaded to the digital twin layer via a communication network.

[0053] Step Two: Construction and Synchronization of Wide-Area Digital Twins for Transmission Lines

[0054] ① Model Construction: In the digital twin layer, based on the power grid GIS, equipment parameters, and physical laws, a high-fidelity, multi-physics coupled digital mirror model is established, covering all transmission lines, substations, generator sets, and loads of the target wide-area power grid. This model can not only reflect steady-state electrical characteristics but also simulate electromagnetic transients, thermal dynamics, and other processes.

[0055] ② Real-time synchronization: Using the real-time data uploaded in step one, the operating status of the digital twin model is dynamically calibrated through technologies such as state estimation and parameter identification, so that it is synchronized with the physical power grid at the millisecond level and truly reflects the current operating condition of the power grid;

[0056] Step 3: Quantum Modeling and Transformation of the Scheduling Problem

[0057] In the quantum-classical hybrid computing layer, scheduling objectives and constraints are received from the application service layer. Through a transformation algorithm, the above-mentioned classical power grid wide-area optimization scheduling problem is mapped or approximately equivalent to a quadratic unconstrained binary optimization QUBO model or the Ising model.

[0058] Step 4: Quantum-Classical Hybrid Solution

[0059] The constructed QUBO model is input into a quantum computing device (such as a quantum annealing machine or a noisy medium-scale quantum-NISQ computer).

[0060] Quantum computing devices utilize quantum parallelism for rapid sampling and search for the lowest energy state, outputting one or more sets of optimized solutions (corresponding to the values ​​of qubits) in a short time.

[0061] Step 5: Scheme Demonstration and Decision Making

[0062] The optimized scheduling scheme and anticipated operation set obtained in step four are injected into the real-time synchronized digital twin model in step two for advanced simulation and N-k safety verification.

[0063] The digital twin quickly calculates the state changes of the power grid in the next scheduling cycle; if the calculation results meet all security constraints, the scheme is confirmed and an optimized scheduling instruction is formed; if a limit violation is found, the new constraints or penalty terms are fed back to step three for iterative optimization until a safe and feasible optimal scheme is obtained and an optimized scheduling instruction is formed.

[0064] Step Six: Issuance and Execution of Commands

[0065] The finalized optimized scheduling instructions are sent to the execution mechanism at the physical power grid layer through a secure channel to complete closed-loop control.

[0066] The intelligent sensor in step one includes a PMU and an intelligent terminal; the electrical quantities of the power transmission line collected include I, V, P, and Q; the equipment status includes switches and transformer taps; and the environmental data includes temperature and wind speed.

[0067] The application service layer in step three receives scheduling objectives including minimizing network loss, minimizing power generation cost, and maximizing safety margin. The constraints include line power flow limits, voltage upper and lower limits, and generator output range. The power grid wide-area optimization scheduling problem typically involves a mixed integer programming problem that is nonlinear and contains discrete variables.

[0068] The transformation algorithm in step three discretizes continuous variables (such as generator active power output) into a combination of multiple qubits, and encodes the objective function and constraints (including equality constraints and inequality constraints) into linear and quadratic terms in the Hamiltonian of the QUBO model.

[0069] The optimized scheduling instructions in step six include AGC setpoints and switch operation commands, and the actuators include the power plant control system and the substation automation system.

[0070] In this invention, the classical computing coprocessor is responsible for: 1) decoding the original solution output by quantum computing and converting it back into power grid scheduling variables (such as unit start-up and shutdown status, output value); 2) performing feasibility verification and local fine optimization (for example, using the classical QP algorithm to quickly solve the continuous power flow optimization based on the excellent discrete variable combination provided by the quantum solution), so as to make up for the lack of precision and scale of current quantum hardware.

[0071] A wide-area optimization scheduling device for transmission lines based on quantum computing and digital twins comprises a data acquisition and communication module, a digital twin modeling and synchronization module, a problem modeling and transformation module, a quantum-classical hybrid computing module, a scheme verification and decision-making module, and an instruction generation and issuance module. As a hybrid enhanced intelligent scheduling system, it realizes real-time data interaction between layers through a high-speed communication network, and is used to implement the aforementioned wide-area optimization scheduling method for transmission lines based on quantum computing and digital twins.

[0072] Specifically, it includes:

[0073] Data Acquisition and Communication Module: Configured for acquiring real-time data from the physical power grid; Digital Twin Modeling and Synchronization Module: Configured for building and updating the power grid digital twin model; Problem Modeling and Transformation Module: Configured for converting scheduling requirements into quantum optimization models; Quantum-Classical Hybrid Computing Module: Connected to quantum computing devices, responsible for allocating scheduling tasks, receiving and decoding quantum computing results, and classical post-processing; Scheme Verification and Decision Module: Configured for verifying optimization schemes in the digital twin environment; Instruction Generation and Issuance Module: Configured for generating executable optimized scheduling instructions.

[0074] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements all the steps of the aforementioned wide-area optimal scheduling of transmission lines based on quantum computing and digital twins.

[0075] In an embodiment, the conversion algorithm described in this invention is as follows:

[0076] I. Problem Definition (Classic Wide-Area Optimal Scheduling Model)

[0077] 1.1 Objective Function (Minimize Network Loss)

[0078]

[0079] Where: E: collection of transmission lines; G ij : Line i__j conductance; V i : Voltage amplitude at node i; V j : Voltage amplitude at node j; δ ij = δi__δj :Node voltage phase angle difference.

[0080] 1.2 Constraints

[0081] (1) Power balance constraint (equality constraint)

[0082]

[0083] Where: P Gi Generator i has active power output; PDi : Active load of node i; B ij Line i__j susceptance; N i : The set of adjacent nodes of node i.

[0084] (2) Variable boundary constraints (inequality constraints)

[0085] P Gmin,i ≤ P Gi ≤ P Gmax,i V min,i ≤ V i ≤ V max,i ;δ min,ij ≤ δ ij ≤ δ max,ij .

[0086] (3) Line power flow constraints (inequality constraints)

[0087] |P ij |≤ P ij,max

[0088] P ij = G ij V i V j cosδ ij + B ij V i V j sinδ ij Line i__j has active power flow.

[0089] II. Core Steps and Formulas of the Conversion Algorithm

[0090] Step 1: Discretization of continuous variables and encoding of qubits

[0091] All continuous variables (P) Gi V i δ ij The variables are discrete into a combination of binary variables, and each variable is encoded with m qubits (the precision is determined by m).

[0092] 1.1 Generator active power output P Gi coding

[0093]

[0094] in:

[0095] x i,k ∈ 0,1: Encoding P Gi The kth binary variable (corresponding to a qubit);

[0096] : It's a long walk away;

[0097] m: Number of encoding bits (recommended m = 4~6, balancing accuracy and computing power).

[0098] 1.2 Node voltage V i coding

[0099]

[0100] in:

[0101] y i,k ∈ 0,1: Encoding V i The kth binary variable;

[0102] .

[0103] 1.3 Voltage phase angle difference δ ij coding

[0104]

[0105] in:

[0106] z ij,k ∈ 0,1: Encoding δ ij The kth binary variable;

[0107] .

[0108] 1.4 Unified Representation of Binary Variables

[0109] Define the total set of binary variables: 8 = x i,k ,y i,k ,z ij,k There are a total of n = m . (G + N + E) variables (G: number of generators, N: number of nodes, E: number of lines).

[0110] Step 2: Objective function (network loss) QUBO mapping

[0111] Transform the original objective function J (network loss) into a QUBO quadratic form: , where h a J is the coefficient of the linear term. ab The coefficient of the quadratic term, s a Let s be the a-th binary decision variable (0-1 variable). b Let b be the b-th binary decision variable (0-1 variable).

[0112] 2.1 Network loss function expansion (substituting discretized variables)

[0113]

[0114] Where: y i,k ∈{0,1} : Encoding V i The k-th binary variable; k=0; For each node j in {0,1}, the voltage magnitude Vj is binary encoded / discrete. This represents the l-th binary bit of the voltage encoding of node j.

[0115] 2.2 Nonlinear term approximation (Cosine term Taylor expansion)

[0116]

[0117] Substitute δ ij Discretize the expression, expand it and retain only the quadratic term (as required by the QUBO model), and suppress higher-order terms through weight coefficients.

[0118] 2.3 Final QUBO Form of the Objective Function

[0119]

[0120] Coefficient calculation:

[0121] Linear term h a All containing s a The sum of the coefficients of the linear terms;

[0122] Quadratic term J ab All containing s a s b The sum of the cross term coefficients.

[0123] Step 3: Constraint QUBO Embedding (Penalty Function Method)

[0124] Transform equality / inequality constraints into penalty terms, embed them into the QUBO objective function, and construct an unconstrained optimization problem:

[0125] H QUBO = J QUBO +λ1.C1 (8)+λ2.C2 (8)+λ3.C3 (8)

[0126] λ1,λ2,λ3: Penalty coefficients (must satisfy λ1,λ2,λ3) J QUBO Maximum value (ensuring constraints are satisfied);

[0127] C1, C2, and C3 are penalty terms for power balance, variable boundary conditions, and line power flow constraints, respectively.

[0128] 3.1 Power balance constraint penalty term C1

[0129]

[0130] Substitute P Gi V i δ ij The discretized expression can be expanded into a quadratic form with binary variables:

[0131]

[0132] 3.2 Variable boundary constraint penalty term C2

[0133] Employing "boundary violation penalty" with P Gi For example (V) i δ ij Similarly):

[0134]

[0135] After discretization, it transforms into a quadratic form:

[0136]

[0137] 3.3 Line power flow constraint penalty item C3

[0138]

[0139] After discretization, it transforms into a quadratic form:

[0140] .

[0141] Step 4: QUBO → Ising Model Conversion (Optional)

[0142] The variable in the Ising model is σ a ∈_1,1, and QUBO binary variable x a Transformation relation for ∈0,1: Substituting into the QUBO Hamiltonian, we obtain the Ising model:

[0143]

[0144] Coefficient conversion formula:

[0145] ;

[0146] ;

[0147] (Constant term, does not affect optimization).

[0148] III. Summary of Complete Conversion Formulas

[0149] 1. General Formula for Variable Encoding

[0150]

[0151] 2. QUBO's total Hamiltonian formula

[0152]

[0153] 3. Simplified formula for coefficient calculation (taking the linear term ha of network loss as an example)

[0154]

[0155] Quadratic terms The calculation logic for constraint penalty coefficients c1, a, c1, ab, etc. is consistent.

[0156] IV. Recommendations for Key Parameter Selection

[0157] 1. Number of encoded bits m: 4~6 (precision ≈ 6 when m = 4, precision ≈ 1.5 when m = 6).

[0158] 2. Penalty coefficient λ: λ1=λ2=λ3= 10 3 ~ 10 4 (It needs to be calibrated according to the magnitude of the objective function to ensure that the constraint satisfaction has a higher priority than minimizing network loss).

[0159] 3. Taylor expansion order: The cosine term is retained up to order 4 (balancing accuracy and computational complexity).

[0160] V. Application Instructions

[0161] 1. The H output by this algorithm QUBO The QAOA quantum algorithm can be directly input to solve the problem. The solution is then used to obtain the generator output, voltage, and other scheduling instructions through the "binary to decimal" reverse encoding.

[0162] 2. Nonlinear terms (such as V) i V j cosδ ij The second approximation error can be compensated by increasing the number of bits m or adjusting the penalty coefficient λ;

[0163] 3. Applicable to wide-area dispatching scenarios including generators, loads, and transmission lines, and can be extended to complex power grids including energy storage and new energy sources.

[0164] Explanation of related terms in this invention

[0165] Quantum computing: A new computing paradigm that utilizes the superposition and entanglement properties of qubits for information processing, with the potential for exponential speedup in handling specific complex combinatorial optimization problems.

[0166] Digital twin: A digital model built in virtual space that is fully mapped to and interacts with a physical entity in real time, which can be used to simulate, predict, and optimize the behavior of the physical entity.

[0167] Wide-area optimal dispatch: refers to making optimal decisions on the system operation status by comprehensively considering multiple constraints such as power generation, power transmission, and load within a cross-regional and large-scale power grid.

[0168] QUBO model: a quadratic unconstrained binary optimization model, is a mathematical model that can be used to describe many combinatorial optimization problems and is a typical input form for direct solution by current quantum annealing machines.

[0169] Quantum annealing: a quantum computing technique that uses the quantum tunneling effect to find the global optimal solution, and is particularly suitable for solving combinatorial optimization problems.

[0170] Alternative solution to the invention:

[0171] I. Alternatives to Quantum Computing Modules

[0172] (I) Classic High-Performance Computing Cluster Acceleration Solution

[0173] When quantum computing hardware cannot yet meet the needs of practical applications, classical high-performance computing clusters can be used as an alternative. By constructing a distributed computing cluster consisting of dozens to hundreds of high-performance servers, and combining it with parallel computing frameworks (such as Spark and MPI), traditional optimization algorithms can be parallelized. For example, when solving large-scale optimal power flow problems, the power grid can be decomposed into regions, and the computational tasks of each sub-region can be distributed to different computing nodes for parallel processing. Finally, a coordination algorithm is used to aggregate the global optimal solution. This approach relies on mature classical computing technology and can significantly improve computational efficiency under existing hardware conditions. Although it cannot achieve the theoretical speedup of quantum computing, it can meet the real-time requirements of most current power grid dispatching scenarios.

[0174] (II) Integration of Heuristic Algorithms and Machine Learning

[0175] For complex combinatorial optimization problems in power grid dispatching, an alternative approach combining heuristic algorithms and machine learning can be adopted. On the one hand, the global search capabilities of heuristic algorithms such as genetic algorithms and particle swarm optimization can be used to quickly generate a set of candidate dispatching schemes. On the other hand, a prediction model based on deep learning can be constructed to quickly evaluate and rank the performance of candidate schemes, selecting the most promising schemes for fine-tuning. For example, Long Short-Term Memory (LSTM) networks can be used to learn from historical dispatching data to predict key indicators such as network losses and voltage stability under different schemes, thereby reducing the number of iterations of heuristic algorithms and improving solution efficiency. This approach does not rely on quantum computing hardware and has strong flexibility and adaptability, enabling dynamic adjustments based on the actual operation of the power grid.

[0176] II. Alternatives to Digital Twin Modules

[0177] (I) Multi-dimensional simulation model integration scheme

[0178] If the construction cost of digital twin technology is too high or the technical difficulty is too great, a multi-dimensional simulation model integration scheme can be used as an alternative. By integrating modules such as power flow calculation, transient stability analysis, and electromagnetic transient simulation from power system analysis software (e.g., PSASP, PSS / E), a comprehensive simulation platform covering the steady-state, dynamic, and transient characteristics of the power grid can be built. This platform can quickly simulate and evaluate different dispatching schemes based on real-time collected power grid operation data, providing technical support for dispatching decisions. Compared with digital twin models, the multi-dimensional simulation model integration scheme, although lacking real-time mapping and interaction capabilities, can meet the simulation analysis needs of power grid dispatching to a certain extent, and has a lower technical threshold and implementation cost.

[0179] (II) Data-driven virtual scene construction scheme

[0180] Based on big data analytics and data mining technologies, a data-driven virtual scenario is constructed as an alternative to the digital twin model. By analyzing and mining massive amounts of historical power grid operation data, real-time monitoring data, and external environmental data (such as meteorological data and load data), the inherent patterns and characteristics of power grid operation are extracted to construct a virtual scenario that reflects the actual operating state of the power grid. For example, clustering algorithms are used to classify historical load data, constructing virtual scenarios under different load levels; association rule mining techniques are used to analyze the relationship between meteorological data and power grid operating parameters, enabling the prediction and assessment of the power grid's operating state. This solution does not rely on a precise physical model, can quickly adapt to changes and developments in the power grid, and provides real-time and accurate reference data for dispatching decisions.

[0181] III. Alternatives to Quantum-Classical Hybrid Computing Architectures

[0182] (a) Layered and progressive computing architecture

[0183] A hierarchical, progressive computing architecture replaces the quantum-classical hybrid computing architecture. The power grid scheduling problem is decomposed into multiple levels of sub-problems, solved progressively from global to local. At the top level, classical optimization algorithms are used to macroscopically optimize the overall operation of the power grid, determining the generation plans and transmission power allocation for each region. In the middle level, more refined optimization algorithms are employed to solve problems related to unit combination and reactive power optimization, tailored to the specific conditions of each region. At the bottom level, real-time control technology is used to adjust and control the power grid's operation in real time. This architecture, through its hierarchical and progressive approach, decomposes the complex scheduling problem into multiple simpler sub-problems, reducing computational complexity, improving solution efficiency, and avoiding dependence on quantum computing hardware.

[0184] (II) Cloud-Edge Collaborative Computing Architecture

[0185] A cloud-edge collaborative computing architecture is constructed to fully leverage the powerful computing capabilities of cloud computing and the real-time processing capabilities of edge computing. Large-scale optimization tasks in power grid dispatching are deployed in the cloud, utilizing the distributed computing resources of the cloud platform for rapid solutions. Real-time data processing, status monitoring, and control tasks are deployed on edge nodes, enabling real-time perception and response to the power grid's operational status. Through cloud-edge collaboration, the rational allocation of computing resources and efficient task execution are achieved, providing low-latency, highly reliable computing support for power grid dispatching. This architecture does not rely on quantum computing technology and possesses excellent scalability and flexibility, adapting to the continuous expansion of the power grid and the evolving business needs.

[0186] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for wide-area optimal scheduling of power transmission lines based on quantum computing and digital twins, characterized in that: A high-fidelity digital twin model of the physical entity of the transmission line is constructed to synchronize the physical power grid status in real time; at the same time, the complex wide-area power grid optimization scheduling problem is modeled into an optimization model suitable for quantum computer solution. Leveraging the powerful parallel computing capabilities of quantum computing, scheduling schemes are rapidly solved and simulated in a digital twin environment; finally, the optimized scheduling instructions are sent to the physical power grid for execution, achieving globally optimal and dynamically adaptive intelligent scheduling.

2. The method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins according to claim 1, characterized in that: The complex power grid wide-area optimization scheduling problem includes power flow optimization, economic scheduling, and security verification; the optimization model is a quadratic unconstrained binary optimization-QUBO model, and the quantum computing is a quantum annealing or gate circuit model.

3. The method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins according to claim 2, characterized in that... Includes the following steps: Step 1: Real-time sensing and data upload of the physical power grid Smart sensors distributed throughout the power grid collect electrical quantities, equipment status, environmental data, and power prediction data of new energy power plants in real time, and upload the real-time data to the digital twin layer through the communication network. Step Two: Construction and Synchronization of Wide-Area Digital Twins for Transmission Lines ① Model Construction: In the digital twin layer, based on the power grid GIS, equipment parameters and physical laws, a high-fidelity, multi-physics coupled digital mirror model is established, covering all transmission lines, substations, generator sets and loads of the target wide-area power grid; this model can reflect steady-state electrical characteristics and simulate electromagnetic transient and thermal dynamic processes. ② Real-time synchronization: Using the real-time data uploaded in step one, the operating status of the digital twin model is dynamically calibrated through state estimation and parameter identification, so that it is synchronized with the physical power grid at the millisecond level and truly reflects the current operating condition of the power grid; Step 3: Quantum Modeling and Transformation of the Scheduling Problem In the quantum-classical hybrid computing layer, scheduling objectives and constraints are received from the application service layer; through a transformation algorithm, the above-mentioned power grid wide-area optimization scheduling problem is mapped or approximately equivalent to a quadratic unconstrained binary optimization QUBO model or the Ising model. Step 4: Quantum-Classical Hybrid Solution The constructed QUBO model is input into the quantum computing device; Quantum computing devices utilize quantum parallelism for rapid sampling and search for the lowest energy state, outputting one or more sets of optimized solutions in a short time. Step 5: Scheme Demonstration and Decision Making The optimized scheduling scheme and anticipated operation set obtained in step four are injected into the real-time synchronized digital twin model in step two for advanced simulation and N-k safety verification. The digital twin quickly calculates the state changes of the power grid in the next scheduling cycle; if the calculation results meet all security constraints, the scheme is confirmed and an optimized scheduling instruction is formed; if a limit violation is found, the new constraints or penalty terms are fed back to step three for iterative optimization until a safe and feasible optimal scheme is obtained and an optimized scheduling instruction is formed. Step Six: Issuance and Execution of Commands The finalized optimized scheduling instructions are sent to the execution mechanism at the physical power grid layer through a secure channel to complete closed-loop control.

4. The method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins according to claim 3, characterized in that: The application service layer in step three receives scheduling objectives including minimizing network loss, minimizing generation cost, and maximizing safety margin. The constraints include line power flow limits, voltage upper and lower limits, and generator output range. The power grid wide-area optimization scheduling problem typically involves a mixed integer programming problem that is nonlinear and contains discrete variables.

5. The method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins according to claim 3, characterized in that: The transformation algorithm in step three discretizes the continuous variable into a combination of multiple qubits and encodes the objective function and constraints into linear and quadratic terms in the Hamiltonian of the QUBO model.

6. The method for wide-area optimal scheduling of transmission lines based on quantum computing and digital twins according to claim 3, characterized in that: The optimized scheduling instructions in step six include AGC setpoints and switch operation commands, and the actuators include the power plant control system and the substation automation system.

7. A wide-area optimized scheduling device for power transmission lines based on quantum computing and digital twins, characterized in that: Composed of a data acquisition and communication module, a digital twin modeling and synchronization module, a problem modeling and transformation module, a quantum-classical hybrid computing module, a scheme verification and decision-making module, and an instruction generation and issuance module, this system serves as a hybrid enhanced intelligent scheduling system. It enables real-time data interaction between different layers through a high-speed communication network and is used to implement the wide-area optimized scheduling method for transmission lines based on quantum computing and digital twins as described in any one of claims 1-6.

8. A wide-area optimized scheduling device for transmission lines based on quantum computing and digital twins according to claim 7, characterized in that... Specifically, it includes: Data Acquisition and Communication Module: Configured for acquiring real-time data from the physical power grid; Digital Twin Modeling and Synchronization Module: Configured for building and updating the power grid digital twin model; Problem Modeling and Transformation Module: Configured for converting scheduling requirements into quantum optimization models; Quantum-Classical Hybrid Computing Module: Connected to quantum computing devices, responsible for allocating scheduling tasks, receiving and decoding quantum computing results, and classical post-processing; Scheme Verification and Decision Module: Configured for verifying optimization schemes in the digital twin environment; Instruction Generation and Issuance Module: Configured for generating executable optimized scheduling instructions.

9. A computer-readable storage medium, characterized in that... It stores a computer program, which, when executed by a processor, implements all the steps of a wide-area optimized scheduling of transmission lines based on quantum computing and digital twins as described in any one of claims 1-6.

Images