Source network load storage cooperative operation method and system based on multi-factor coupling

Abstract

The invention relates to the technical field of power systems, and discloses a source-network-load-storage cooperative operation method and system based on multi-factor coupling, and the method comprises the steps: constructing a system state vector; modeling a multi-factor coupled space-time potential field; potential field wave evolution prediction; calculating total cooperative control force; and carrying out distributed resource closed-loop control. According to the method and the system, the space-time influence of propagation and evolution of disturbance in the power grid can be predicted in real time, multiple resources of the whole network are organized to perform prospective and cooperative active defense, distributed resources can be optimally utilized to quickly suppress disturbance, and the dynamic safety margin is improved.

Description

Technical Field

[0001] This invention relates to the field of power system technology, specifically to a method and system for coordinated operation of power generation, grid, load and storage based on multi-factor coupling, which is particularly suitable for coordinated operation control of power generation, grid, load and storage systems with a high proportion of renewable energy. Background Technology

[0002] With the increasing penetration of renewable energy sources such as wind and solar power into the power grid, the operating characteristics of the power system have undergone profound changes. Ensuring the safe, stable, and economical operation of the system under conditions of intermittent and fluctuating renewable energy output and increased load uncertainty, and achieving deep synergy and efficient interaction among various flexible resources such as power generation, grid, load, and storage, has become a core technological challenge in this field.

[0003] In existing technologies, the coordinated operation of power systems mainly relies on a centralized control architecture based on "prediction-optimization". A typical model involves the power grid dispatch and control center collecting measurement data from the entire network and, based on short-term forecasts of renewable energy output and load over a future period, calculating the output plans or control settings for each controllable unit, such as generating units and energy storage power stations, using optimization algorithms such as optimal power flow and economic dispatch, with the goal of system operation economy or safety. The commands are then issued to each substation for execution. For rapid control requirements such as frequency stability, the system relies on preset strategies such as local droop control in each unit, combined with regional-level automatic generation control (AGC).

[0004] However, the aforementioned centralized control model based on "prediction-planning-execution" suffers from inherent time lag and insufficient foresight when dealing with high-frequency, highly uncertain disturbances brought about by high-proportion renewable energy systems. When unpredictable disturbance events occur in the system (such as a sudden drop in renewable energy output or line faults), this control model mainly relies on the detection and correction of post-event deviations. It lacks a mechanism that can predict the spatiotemporal impact of disturbances propagating and evolving in the power grid in real time and organize multi-resource resources across the entire network for proactive and coordinated defense. This results in a system response that is usually passive and delayed, making it difficult to optimally utilize distributed resources to quickly suppress disturbances and affecting the system's dynamic security margin.

[0005] Therefore, proposing a source-grid-load-storage coordinated operation method and system based on multi-factor coupling has become an urgent problem to be solved. Summary of the Invention

[0006] In view of this, the present invention provides a source-grid-load-storage coordinated operation method based on multi-factor coupling to solve the problem that the existing technology cannot predict the spatiotemporal propagation impact of disturbances and organize resources for forward-looking coordinated defense, resulting in a passive and delayed response when dealing with emergencies.

[0007] To achieve the above objectives, this invention provides a source-grid-load-storage coordinated operation method based on multi-factor coupling, comprising the following steps:

[0008] Real-time acquisition and preprocessing of multi-source heterogeneous data from the source-grid-load-storage system yields the real-time system state vector;

[0009] Based on the real-time system state vector, multiple basic sub-potential field functions are defined, and a dynamic potential field coupling tensor is used to nonlinearly couple the basic sub-potential field functions to form a multi-factor coupled spatiotemporal potential field. When the tensor elements in the potential field coupling tensor exceed a preset threshold, the tensor elements are nonlinearly enhanced.

[0010] The disturbance sources in the system are identified and the initial potential energy disturbance is quantified. Then, the future spatiotemporal distribution of the potential wave is predicted using the potential wave propagation model.

[0011] The reactive control force is calculated based on the real-time gradient of the spatiotemporal potential field, and the forward control force is calculated based on the gradient expectation of the future spatiotemporal distribution of the potential field wave. The reactive control force and the forward control force are then combined into a total cooperative control force.

[0012] The overall collaborative control force is distributed to the distributed resource side. The distributed resource side makes local optimization decisions based on the overall collaborative control force and its own operational constraints, and feeds back the execution results to the system to achieve closed-loop control of distributed resources.

[0013] Preferably, the multi-source heterogeneous data includes wide-area measurement system data, SCADA / EMS system data, meteorological forecast data, and electricity market transaction data.

[0014] Further preferably, the preprocessing includes data cleaning, outlier removal, and data standardization.

[0015] Further preferably, the elements of the real-time system state vector include system frequency, key node voltage, active power output of the new energy station, total load power of the load aggregator, and state of charge of the energy storage unit.

[0016] Further optimization reveals that the defined fundamental subpotential functions include:

[0017] Equilibrium potential field function: used to quantify the risk of power supply and demand imbalance in the overall system;

[0018] Safety potential function: used to quantify how well the system's operating state approximates the safety constraints;

[0019] The reserve potential field function is used to quantify the risk of a system's reserve capacity adequacy in response to future uncertainties.

[0020] Flexibility potential field function: used to quantify the state health risk of flexible resources such as energy storage and demand-side response.

[0021] Further preferred, the potential wave propagation model is a discrete wave equation based on the power grid diagram structure. The parameters in the equation include attenuation coefficient and propagation coefficient. The attenuation coefficient of node n is used to characterize the local dissipation rate of potential energy at node n, and the propagation coefficient of node n is used to characterize the rate and intensity of potential energy propagating from neighboring node j to node n.

[0022] Further optimization involves incorporating a look-ahead time domain and a time discount factor into the calculation of the forward control force. The length of the look-ahead time domain is set according to the type of disturbance, and the time discount factor decreases as the time interval increases.

[0023] In a further preferred embodiment, after receiving the total collaborative control force, the distributed resource uses it as the guiding term of the objective function of the local optimization problem to solve for the optimal control action that satisfies all its own physical and economic constraints.

[0024] This invention also provides a source-grid-load-storage coordinated operation system based on multi-factor coupling, comprising:

[0025] The system state vector construction module is used to collect multi-source heterogeneous data from the source-grid-load-storage system in real time and preprocess it to obtain the real-time system state vector.

[0026] The spatiotemporal potential field coupling module is used to define multiple basic sub-potential field functions based on the real-time system state vector, and to nonlinearly couple the basic sub-potential field functions using a dynamic potential field coupling tensor to form a multi-factor coupled spatiotemporal potential field. The tensor elements in the potential field coupling tensor are used to perform nonlinear enhancement when the system indicators related to the tensor elements exceed a preset threshold.

[0027] The potential wave future spatiotemporal distribution prediction module is used to identify disturbance sources in the system and quantify the initial potential energy disturbance. Then, the potential wave propagation model is used to predict the future spatiotemporal distribution of the potential wave.

[0028] The overall collaborative control force calculation module is used to calculate the reactive control force based on the real-time gradient of the spatiotemporal potential field, calculate the forward control force based on the gradient expectation of the future spatiotemporal distribution of the potential field wave, and combine the reactive control force and the forward control force into the overall collaborative control force.

[0029] The distributed resource closed-loop control module is used to distribute the total collaborative control force to the distributed resource side. The distributed resource side makes local optimization decisions based on the total collaborative control force and its own operating constraints, and feeds back the execution results to the system to realize the closed-loop control of distributed resources.

[0030] Preferably, the total collaborative control force calculation module includes a forward control force calculation module. The calculation of the forward control force introduces a forward time domain and a time discount factor. The length of the forward time domain is set according to the disturbance type, and the time discount factor decreases as the time interval increases.

[0031] The source-grid-load-storage coordinated operation method and system based on multi-factor coupling provided by this invention have the following beneficial effects:

[0032] 1. This invention utilizes a multi-source heterogeneous data fusion and spatiotemporal potential field unified quantization algorithm to form a closed loop, integrating system state vector construction, multi-factor coupled potential field modeling, collaborative control force calculation, and distributed local optimization. Specifically, the dynamic reconstruction mechanism of the potential field coupling tensor internalizes external factors such as meteorology and market conditions into dynamic adjustments to the system's operational objectives; the issued total collaborative control force guides each resource entity to autonomously seek optimization while satisfying its own constraints. This closed-loop linkage improves the accuracy of system supply and demand balance and the capacity for renewable energy absorption under full-factor coupling.

[0033] 2. This invention quantifies physical disturbances into initial potential energy by linking algorithms for potential wave evolution prediction and forward-looking control force calculation, and predicts its spatiotemporal distribution based on the power grid topology. The collaborative control force calculation module then generates forward-looking control components to guide system resources in pre-responding to impending risk shocks. This mechanism realizes a shift in the system control paradigm from passive post-event response to proactive pre-disturbance defense, significantly enhancing the system's transient stability under scenarios of sudden changes in new energy output or large load fluctuations.

[0034] 3. This invention transforms system-level regulation requirements into local optimization objectives for each resource by linking global guidance of collaborative control with algorithms for distributed resource local optimization. The resource-side controller autonomously solves for optimal control actions after calculating its own operational constraints (such as energy storage SOC and load cost), achieving control decoupling and efficient execution. Simultaneously, based on operational status feedback, the potential field coupling tensor can be reconstructed, ensuring the robustness of the control strategy to changing system characteristics and improving the scalability and resource utilization efficiency of the system control. Attached Figure Description

[0035] Figure 1 The flowchart of the source-grid-load-storage coordinated operation method based on multi-factor coupling provided by the present invention. Detailed Implementation

[0036] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] like Figure 1 As shown, the source-grid-load-storage coordinated operation method based on multi-factor coupling provided by the present invention includes the following steps:

[0038] S1: Collect multi-source heterogeneous data from the source-grid-load-storage system in real time and preprocess it to obtain a standardized real-time system state vector that can comprehensively describe the system state.

[0039] The system collects multi-dimensional operational data from multiple data sources, including: 1) Wide Area Measurement System (WAMS), which provides high-frequency, time-stamped system phasor data, such as node voltage phasors and branch current phasors; 2) SCADA / EMS system, which provides routine grid steady-state operation data, such as generator output, load power, and switch status; 3) Meteorological forecasting platform, which provides meteorological information related to renewable energy output, such as wind speed, wind direction, and solar irradiance; and 4) Electricity market trading platform, which provides economic data such as market clearing prices and ancillary service dispatch.

[0040] The collected raw data needs to be preprocessed to ensure its quality and applicability. The preprocessing process includes: data cleaning, which handles missing values ​​(e.g., through time series interpolation or mean filling) and resolves data conflicts; outlier removal, using the statistical method 3σ criterion to identify and remove data points that exceed the normal fluctuation range; and data standardization, which normalizes data with different physical units and dimensions to a unified interval (e.g., [0,1] or [-1,1]) to eliminate the impact of dimensional differences on subsequent model calculations.

[0041] For example, for a certain data feature Its standardized value It can be calculated using the min-max normalization method:

[0042] ;

[0043] in, and These are the maximum and minimum values ​​of the data feature over a historical period. After preprocessing, the key state information is integrated into a comprehensive representation of the system at time [time value missing]. Real-time system state vector in operation This vector is a high-dimensional vector, and its constituent elements include at least: system frequency. Voltage amplitude at each key node The active power contributions of each new energy power station Total load power of each load aggregator and the state of charge of each energy storage unit. Its mathematical form can be expressed as:

[0044] ;

[0045] in, These represent the sets of nodes, renewable energy power plants, load aggregators, and energy storage units in the system, respectively. This state vector The update cycle is consistent with the data acquisition frequency, which is used to provide real-time, high-quality input for subsequent spatiotemporal potential field modeling;

[0046] S2: Based on the real-time system state vector, multiple basic sub-potential field functions are defined to characterize the system's operational risks from different dimensions. A dynamic potential field coupling tensor is used to nonlinearly couple the basic sub-potential field functions to form a spatiotemporal potential field that can comprehensively reflect the influence of multiple factors. When the tensor elements in the potential field coupling tensor exceed a preset threshold, the tensor elements are nonlinearly enhanced.

[0047] Among them, four fundamental sub-potential field functions are defined, each characterizing the operational risk of the system from different dimensions:

[0048] 1. Equilibrium potential function This is used to quantify the overall power supply and demand imbalance risk of the system. When the total power generation of the system equals the total load (including network losses), the potential energy is zero; the further away from the equilibrium point, the greater the potential energy becomes, increasing quadratically. Its expression is:

[0049] ;

[0050] in, It is a generator Those who have made contributions It is a load active power, It is composed of state vectors Calculated network loss, These are the collections of generators and loads;

[0051] 2. Safety potential field function This is used to quantify how closely the system's operating state approximates safety constraints. When parameters such as line power flow or node voltage approach their safety limits, the potential energy increases sharply, forming a "repulsive barrier." Its expression is:

[0052] ;

[0053] in, It is a side road The trend It is a side road The transmission limit; It is a node voltage, It is a node The upper and lower limits of voltage, It is a node The average value of the voltage limit; These are weighting coefficients. It is an integer greater than 1, used to ensure the steepness of the potential energy as it approaches the limit; These are sets of branches and nodes, respectively.

[0054] 3. Backup potential function This is used to quantify the risk of a system's reserve capacity adequacy in response to future uncertainties. When available reserves are lower than demand, the potential energy increases. Its expression is:

[0055] ;

[0056] in, This is the total reserve capacity currently required by the system. It is a backup resource The available backup capacity under the current conditions. It is a collection of backup resources;

[0057] 4. Flexible potential function This is used to quantify the health risk of flexible resources such as energy storage and demand-side response, preventing overuse of resources that could lead to a loss of continuous regulation capabilities. Its expression is:

[0058] ;

[0059] in, It is an energy storage unit The state of charge, It is an energy storage unit The optimal state of charge; Demand-side resources Fatigue index; These are weighting coefficients; These are a combination of energy storage units and demand-side resources.

[0060] To achieve the dynamic coupling of the aforementioned multiple factors, a potential coupling tensor is introduced. This tensor is a The symmetric matrix. For any node, its subpotential field can be formed as a vector. Then the comprehensive potential energy of that node Calculated using the following quadratic form:

[0061] ;

[0062] Among them, tensor diagonal elements Characterized the first The baseline weights of the sub-potential field, off-diagonal elements Characterized the first and the The coupling strength between the individual potential fields. This tensor possesses dynamic reconstruction capabilities. When key system indicators, such as frequency deviation... Or voltage urgency index When the threshold is exceeded, the tensor will be reconstructed using a preset nonlinear mapping function. For example, when... At that time, the tensor elements related to the equilibrium potential field and the backup potential field Nonlinear enhancement will be achieved through the following S-shaped function:

[0063] ;

[0064] in, It is the benchmark weight. It is the maximum gain. It is the growth slope. It is a trigger threshold. This mechanism enables the system to adaptively adjust the risk dimensions it focuses on under different operating states, achieving dynamic coordination of multiple objectives;

[0065] S3: Identify the disturbance source in the system and quantify the initial potential energy disturbance. Then, use the potential wave propagation model to predict the future spatiotemporal distribution of the potential wave.

[0066] Specifically: identify disturbances in the system and convert them into "wave sources" in the potential field. Then, based on the physical characteristics of the power grid, predict the spatiotemporal propagation process of the "potential field wave" in the future, thereby providing a basis for forward-looking control.

[0067] The first step is the identification and quantification of the disturbance source. The system monitors the state vector at high frequency. The rate of change is used to identify disturbance events. For example, when a large change in the output of generator i is detected within a short period of time... At that time, the event was identified as a source of disturbance. Subsequently, the deviation of this physical quantity was quantified as an initial potential energy disturbance. For example, at the node where the disturbance occurred... Its initial potential energy perturbation It can be quantified as:

[0068] ;

[0069] in, It is the moment when the disturbance occurs. It is a transformation coefficient related to the event type, used to unify different physical disturbances (such as unit failure, line failure, load change) into the potential energy dimension;

[0070] The second step is tuning the propagation model parameters. The propagation of the potential field wave follows a discrete wave equation based on a power grid diagram structure. Key parameters in this equation include the attenuation coefficient. and propagation coefficient It needs to be precisely tuned. Attenuation coefficient. Characterizes the nodes The local dissipation rate of potential energy at a given node is mainly related to the load damping characteristics and local regulation resources (such as the governor of the generator set), and its value can be calculated and calibrated offline based on physical parameters such as line impedance. Propagation coefficient This represents the potential energy from neighboring nodes. propagation to nodes The rate and strength of the signal are positively correlated with the tightness of the electrical connection between the two nodes. They are usually adjusted online adaptively based on the admittance value of the line between the two points to adapt to real-time changes in the power grid topology.

[0071] The third step is to predict the future spatiotemporal distribution of the potential wave. Using a propagation model with tuned parameters, the evolution of the potential wave is predicted through high-frequency iterative solutions. Its discrete-time propagation equation is:

[0072] ;

[0073] in, It is a node At any moment Total potential energy, It is with nodes The set of connected neighbor nodes, Time at the node Newly generated disturbance sources. To meet real-time requirements, this step can be performed using a surrogate model based on a graph neural network (GNN) or a fast numerical solver. The GNN surrogate model, trained offline, learns the dynamic characteristics of the propagation equation and can directly output the potential field distribution for multiple future time steps within nanoseconds. The iteration step size is set according to the typical response speed of the system (e.g., inertial response time, first-order frequency modulation response time) to ensure the effectiveness of the prediction. Finally, the output of this step is a sequence of predicted potential energy values ​​for all nodes in the entire network over a future period (e.g., the next 5-10 seconds). This provides input for the forward-looking control force calculation in S4;

[0074] S4: Calculate the reactive control force based on the real-time gradient of the spatiotemporal potential field, calculate the forward control force based on the gradient expectation of the future spatiotemporal distribution of the potential field wave, and combine the reactive control force and the forward control force into a total cooperative control force.

[0075] This step is used to generate a clear and quantifiable regulatory "driving signal" for each controllable resource in the system, namely the total collaborative control force. This force is dynamically synthesized from two parts: one part is the "reactive control force" that responds to current system risks, and the other part is the "proactive control force" that anticipates and responds to future risks.

[0076] First, calculate the reactive control force. It represents controllable resources (Located at a power grid node) The adjustment action of ) reduces the current moment System local comprehensive potential energy The marginal effect. It is defined as the combined potential energy on the control variable of the resource. The negative gradient, mathematically expressed as:

[0077] ;

[0078] in, That is, resources At the node place, time responsive control force; It is the combined potential energy of the node calculated based on S2; It is a resource The control variables (e.g., the charging and discharging power of energy storage). Due to It is a highly nonlinear function constructed through a potential field coupling tensor, and its gradient is typically calculated using numerical methods. In a preferred embodiment of the present invention, the gradient is calculated using the numerical difference method:

[0079] ;

[0080] in, For control variables A tiny perturbation applied;

[0081] Secondly, calculate forward-looking control. It aims to drive resources to respond in advance to the predicted, upcoming potential wave in S3. This force is obtained by weighted summation of the expected gradients of the predicted potential wave arriving over a future period, and its mathematical expression is:

[0082] ;

[0083] in, It is forward-looking control; It is the future moment predicted by S3. Reaching the node The potential field wave perturbation component; This is the set look-ahead time domain, the length of which is set according to the type of disturbance. For example, for fast-propagating load change disturbances, a shorter time domain can be set. For unit failures with a wide impact, a longer timeframe can be set. ; It is a time discount factor, meaning that the weight of predictions for the more distant future is reduced; for example, it can take the form of exponential decay. ,in, It's the discount rate; It is the expectation operator, used to handle uncertainties in the prediction of potential wave propagation. Gradient expectation. The solution can be obtained by Monte Carlo simulation, which involves simulating the possible propagation paths of the potential field wave through multiple random samplings, calculating the gradient value under each simulation, and finally taking the statistical average value.

[0084] Finally, the reactive control force and the forward-looking control force are linearly superimposed to synthesize the total synergistic control force:

[0085] ;

[0086] in, These are weighting coefficients used to balance the importance of immediate and look-ahead responses; they can adaptively adjust based on the system's operating state. This scalar... Ultimately, it is distributed to various distributed resources, and its positive and negative signs and magnitude represent the direction and intensity of the adjustment of that resource, respectively;

[0087] S5: The overall collaborative control force is sent down to the distributed resource side. The distributed resource side makes local optimization decisions based on the overall collaborative control force and its own operating constraints, and feeds back the execution results to the system to realize closed-loop control of distributed resources.

[0088] Among them, after receiving the total collaborative control force, the distributed resource takes it as the core guiding signal, combines it with its own operating constraints to make local optimization decisions, and feeds the execution results back to the system to form a complete adaptive closed-loop control.

[0089] Among them, overall collaborative control force As an advanced control signal, it can be sent to the local controllers on each distributed resource side via a preset communication protocol that supports high-frequency data transmission (such as IEC 61850 or a customized UDP protocol). This sending cycle is strictly consistent with the update cycle of the system state vector in S1, ensuring the real-time performance of the control commands.

[0090] Upon receiving the total coordinated control force, the local controller of each distributed resource (such as an energy storage system or a controllable load cluster) does not directly use it as the control setpoint. Instead, it uses it as the guiding term of the objective function for its local optimization problem, in order to solve for the optimal control action that satisfies all its own physical and economic constraints. Specifically, the resource... At any moment Solve the following optimization problem to determine its optimal control action. :

[0091] ;

[0092] The constraints are:

[0093] ;

[0094] The objective function aims to maximize the control action. The "responsiveness" to coordinated control while minimizing its adjustment costs. Constraints encompass various physical and economic limitations of the resource. For example, for an energy storage system, the constraints in its constraint model include:

[0095] Climbing speed limit: ,in, It is the maximum power change rate of the energy storage system;

[0096] SOC security scope: ,in These are the upper and lower limits of SOC. It refers to charge / discharge efficiency. It is the rated capacity;

[0097] For an adjustable load cluster, the constraints include:

[0098] Response cost constraints: The incentive contract terms signed with the user must be met;

[0099] User fatigue limit: Avoid making too frequent adjustments to the same user group in a short period of time. The specific model can limit the maximum number of adjustments or the total amount of adjustments per unit time.

[0100] Based on the mathematical form of the objective function and constraints, this local optimization problem can be solved using efficient algorithms such as linear programming (LP) or quadratic programming (QP).

[0101] Finally, the formation and adaptive adjustment of closed-loop control. (Resources) Execute the optimal control action obtained from the solution. Subsequently, its operating state changes. The resource will feed back its actual control actions and the updated local system state (such as actual output power and current SOC of energy storage) to the system's state awareness layer. This feedback information will be used as part of the next round of data acquisition in S1 to construct the system state vector for the next time step. This forms a complete closed-loop control process of "state perception - potential field calculation - control force generation - resource response - state feedback".

[0102] This invention also provides a source-grid-load-storage coordinated operation system based on multi-factor coupling, comprising:

[0103] The system state vector construction module is used to collect multi-source heterogeneous data from the source-grid-load-storage system in real time and preprocess it to obtain the real-time system state vector.

[0104] The spatiotemporal potential field coupling module is used to define multiple basic sub-potential field functions based on the real-time system state vector, and to nonlinearly couple the basic sub-potential field functions using a dynamic potential field coupling tensor to form a multi-factor coupled spatiotemporal potential field. The tensor elements in the potential field coupling tensor are used to perform nonlinear enhancement when the system indicators related to the tensor elements exceed a preset threshold.

[0105] The potential wave future spatiotemporal distribution prediction module is used to identify disturbance sources in the system and quantify the initial potential energy disturbance. Then, the potential wave propagation model is used to predict the future spatiotemporal distribution of the potential wave.

[0106] The overall collaborative control force calculation module is used to calculate the reactive control force based on the real-time gradient of the spatiotemporal potential field, calculate the forward control force based on the gradient expectation of the future spatiotemporal distribution of the potential field wave, and combine the reactive control force and the forward control force into the overall collaborative control force.

[0107] The distributed resource closed-loop control module is used to distribute the total collaborative control force to the distributed resource side. The distributed resource side makes local optimization decisions based on the total collaborative control force and its own operating constraints, and feeds back the execution results to the system to realize the closed-loop control of distributed resources.

[0108] Preferably, the total collaborative control force calculation module includes a forward control force calculation module. The calculation of the forward control force introduces a forward time domain and a time discount factor. The length of the forward time domain is set according to the disturbance type, and the time discount factor decreases as the time interval increases.

[0109] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A source-grid-load-storage coordinated operation method based on multi-factor coupling, characterized in that, Includes the following steps: Real-time acquisition and preprocessing of multi-source heterogeneous data from the source-grid-load-storage system yields the real-time system state vector; Based on the real-time system state vector, multiple basic sub-potential field functions are defined, and a dynamic potential field coupling tensor is used to nonlinearly couple the basic sub-potential field functions to form a multi-factor coupled spatiotemporal potential field. When the tensor elements in the potential field coupling tensor exceed a preset threshold, the tensor elements are nonlinearly enhanced. The disturbance sources in the system are identified and the initial potential energy disturbance is quantified. Then, the future spatiotemporal distribution of the potential wave is predicted using the potential wave propagation model. The reactive control force is calculated based on the real-time gradient of the spatiotemporal potential field, and the forward control force is calculated based on the gradient expectation of the future spatiotemporal distribution of the potential field wave. The reactive control force and the forward control force are then combined into a total cooperative control force. The overall collaborative control force is distributed to the distributed resource side. The distributed resource side makes local optimization decisions based on the overall collaborative control force and its own operational constraints, and feeds back the execution results to the system to achieve closed-loop control of distributed resources.

2. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The multi-source heterogeneous data includes wide-area measurement system data, SCADA / EMS system data, meteorological forecast data, and electricity market transaction data.

3. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The preprocessing includes data cleaning, outlier removal, and data standardization.

4. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The elements of the real-time system state vector include system frequency, key node voltage, active power output of new energy stations, total load power of load aggregators, and state of charge of energy storage units.

5. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The defined fundamental subpotential functions include: Equilibrium potential field function: used to quantify the risk of power supply and demand imbalance in the overall system; Safety potential function: used to quantify how well the system's operating state approximates the safety constraints; The reserve potential field function is used to quantify the risk of a system's reserve capacity adequacy in response to future uncertainties. Flexibility potential field function: used to quantify the state health risk of flexible resources such as energy storage and demand-side response.

6. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The potential wave propagation model is a discrete wave equation based on a power grid diagram structure. The parameters in the equation include attenuation coefficient and propagation coefficient. The attenuation coefficient of node n is used to characterize the local dissipation rate of potential energy at node n, and the propagation coefficient of node n is used to characterize the rate and intensity of potential energy propagating from neighboring node j to node n.

7. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, The calculation of forward control introduces a forward time domain and a time discount factor. The length of the forward time domain is set according to the type of disturbance, and the time discount factor decreases as the time interval increases.

8. The source-grid-load-storage coordinated operation method based on multi-factor coupling according to claim 1, characterized in that, After receiving the overall cooperative control force, the distributed resource uses it as the guiding term of the objective function of the local optimization problem to solve for the optimal control action that satisfies all its physical and economic constraints.

9. A source-grid-load-storage coordinated operation system based on multi-factor coupling, characterized in that, include: The system state vector construction module is used to collect multi-source heterogeneous data from the source-grid-load-storage system in real time and preprocess it to obtain the real-time system state vector. The spatiotemporal potential field coupling module is used to define multiple basic sub-potential field functions based on the real-time system state vector, and to nonlinearly couple the basic sub-potential field functions using a dynamic potential field coupling tensor to form a multi-factor coupled spatiotemporal potential field. The tensor elements in the potential field coupling tensor are used to perform nonlinear enhancement when the system indicators related to the tensor elements exceed a preset threshold. The potential wave future spatiotemporal distribution prediction module is used to identify disturbance sources in the system and quantify the initial potential energy disturbance. Then, the potential wave propagation model is used to predict the future spatiotemporal distribution of the potential wave. The overall collaborative control force calculation module is used to calculate the reactive control force based on the real-time gradient of the spatiotemporal potential field, calculate the forward control force based on the gradient expectation of the future spatiotemporal distribution of the potential field wave, and combine the reactive control force and the forward control force into the overall collaborative control force. The distributed resource closed-loop control module is used to distribute the total collaborative control force to the distributed resource side. The distributed resource side makes local optimization decisions based on the total collaborative control force and its own operating constraints, and feeds back the execution results to the system to realize the closed-loop control of distributed resources.

10. The source-grid-load-storage coordinated operation system based on multi-factor coupling according to claim 9, characterized in that, The total collaborative control force calculation module includes a forward control force calculation module. The calculation of the forward control force introduces a forward time domain and a time discount factor. The length of the forward time domain is set according to the disturbance type, and the time discount factor decreases as the time interval increases.

Images