Safety and stability control method and system after large-scale energy storage power station is connected to power grid

Abstract

The invention provides a safety and stability control method and system after a large-scale energy storage power station accesses a power grid, and the method comprises the steps: determining the type of a stability problem faced by the power grid in response to a preset serious fault of the power grid, and obtaining a real-time operation state, including a charging state or a discharging state, of the large-scale energy storage power station accessed to the power grid; based on a preset differential control rule, a priority control instruction is generated in combination with the combination of the stability problem type and the real-time operation state of the energy storage power station, and the priority control instruction is used for scheduling the energy storage power station to carry out power rapid adjustment; if the adjustment capability of the energy storage power station is not enough to meet the power grid stability requirement, dynamically generating a removal sequence of a grid-connected unit based on a multi-dimensional evaluation index formed by a resource type, an adjustment response speed and an electrical position; and executing a control action according to the priority control instruction or the removal sequence, and performing safety verification through power flow transfer analysis and voltage response prediction before execution so as to avoid an extra power grid stability problem.

Description

Technical Field

[0001] This invention relates to the field of power system control technology, and in particular to a method and system for safe and stable control of a large-scale energy storage power station after it is connected to the power grid. Background Technology

[0002] With the increasing penetration of new energy sources such as wind power and photovoltaics into the power system, the generation-side structure of the power grid is undergoing profound changes. While promoting cleaner energy, this transformation has also brought about significant changes in system operating characteristics, primarily manifested in a decrease in overall system inertia, weakened voltage and frequency support capabilities, and increased power volatility and uncertainty. To address these challenges, large-scale independent energy storage power stations with millisecond-level rapid power regulation capabilities have been widely adopted, becoming a key means to improve grid regulation flexibility and enhance stability support capabilities. However, the bidirectional power characteristics of energy storage power stations in both "charging" and "discharging" states, as well as their rapid and continuous adjustability as a power electronic interface power source / load, are fundamentally different from the operating mechanisms of traditional grid-connected units such as synchronous generators and conventional new energy units. This presents new compatibility issues for existing safety and stability control systems designed with traditional power sources as the control objects.

[0003] The core logic of existing power grid security and stability control systems' generator tripping strategies largely originated from the era of power systems dominated by synchronous power sources. Their typical operating mode is as follows: based on offline simulation calculations, a "strategy table" is generated for a pre-set set of severe faults, primarily targeting the disconnection of traditional generator units, renewable energy plants, or interruptible loads. During online operation, the system matches the pre-set fault type with real-time collected electrical quantities (such as power and voltage) and protection signals, and executes corresponding "switching quantity" control, mainly in the form of "disconnecting the grid-connected switch." This mode reveals several mismatches when dealing with energy storage power stations. Firstly, in terms of the perception of the controlled object, existing strategies typically simplify the grid-connected unit into a single "power source" or "load," while the dynamic operating state (charging or discharging) of an energy storage power station directly determines its impact on the system's power deficit or surplus at the moment of a fault. The control requirements for the same stability problem from the same energy storage station may be diametrically opposed under different states, and existing strategies lack the perception and decision-making response to this key state variable. Secondly, in terms of control methods, traditional "shutdown" is a discrete, either-or "on / off" control. While direct and effective, it is costly and may cause unnecessary power loss or secondary frequency spikes. The continuous and rapid power regulation capabilities of energy storage power stations offer the possibility of more refined and cost-effective "analog" control. However, existing strategy frameworks fail to systematically integrate and optimize "rapid power regulation" as a priority over "shutdown" control option. Thirdly, regarding the optimal selection of control resources, existing strategies prioritize based on relatively static or electrical attribute parameters such as unit type, installed capacity, and electrical distance. This makes it difficult to effectively quantify and incorporate dynamic, economic, and modal correlation indicators unique to new resources like energy storage power stations, such as "regulation speed," "regulation cost (e.g., cycle life loss)," and "coupling strength with specific stability problem modes (e.g., oscillation modes)." This can lead to the inability to select the optimal control combination when multiple resources need to be coordinated.

[0004] Therefore, when large-scale energy storage power stations are connected, simply using the existing generator disconnection strategy framework and treating them merely as a disconnectable "black box" unit will not only fail to fully utilize their technological potential but may also affect control effectiveness due to control strategy mismatch, and even trigger new stability risks. Currently, there is an urgent need for a new security control strategy that can accurately sense the operating status of energy storage, deeply integrate its rapid power regulation capabilities, and achieve optimized decision-making in multi-resource coordination, in order to adapt to the new requirements of grid security and stability control under the high proportion of new energy and power electronic equipment connected. Summary of the Invention

[0005] To address the shortcomings and deficiencies of existing technologies, this invention provides a method and system for safe and stable control of large-scale energy storage power stations connected to the power grid. Upon responding to a severe grid fault, the method first determines the type of stability problem faced by the grid and acquires the real-time charging or discharging operating status of the energy storage power station. Based on preset differentiated control rules, a priority control command is generated by combining the stability problem type and the energy storage operating status to schedule the energy storage power station for rapid power adjustment, fully leveraging its millisecond-level response advantage to achieve "adjustment instead of switching." When the adjustment capacity of the energy storage power station is insufficient to meet the grid stability requirements, a multi-dimensional evaluation index based on resource type, adjustment response speed, and electrical location is dynamically weighted and a comprehensive priority score is calculated to generate a grid-connected unit disconnection sequence. Before executing control actions, power flow changes are predicted using the power transfer distribution factor method, and voltage response is estimated based on voltage-reactive power sensitivity analysis to conduct safety checks and avoid triggering new stability problems. After execution, the weight coefficients of the multi-dimensional evaluation indexes and the parameters in the differentiated control rules can be optimized offline or online based on control effect data to achieve adaptive strategy optimization. Among them, the stability problem types are prioritized using hierarchical discrimination logic, and the differentiated control rules are designed to formulate exclusive control schemes for different stability problem types and energy storage operation status combinations. This effectively adapts to the dual functional attributes of energy storage power stations and solves the problems of single control methods, fixed priority settings, and insufficient adaptability of traditional strategies. While ensuring the safe and stable operation of the power grid, it optimizes control costs and improves operational economy.

[0006] The present invention specifically adopts the following technical solution:

[0007] A method for safe and stable control of a large-scale energy storage power station after its connection to the power grid, comprising:

[0008] In response to pre-set severe faults in the power grid, the type of stability problem faced by the power grid is determined, and the real-time operating status of large-scale energy storage power stations connected to the power grid, including charging or discharging status, is obtained.

[0009] Based on preset differentiated control rules, and combined with the combination of the stability problem type and the real-time operating status of the energy storage power station, priority control instructions are generated. These priority control instructions are used to schedule the energy storage power station to perform rapid power adjustment.

[0010] If the regulation capacity of the energy storage power station is insufficient to meet the grid stability requirements, a grid-connected unit disconnection sequence is dynamically generated based on a multi-dimensional evaluation index consisting of resource type, regulation response speed and electrical location.

[0011] Control actions are executed according to the aforementioned priority control commands or cut-off sequences. Before execution, safety checks are performed through power flow transfer analysis and voltage response prediction to avoid triggering additional power grid stability problems.

[0012] Furthermore, the stability problem types include transient power angle stability, transient voltage stability, thermal stability, and dynamic stability, and a hierarchical discrimination logic is used to determine the priority, with the specific priority order being transient power angle stability > transient voltage stability > thermal stability > dynamic stability.

[0013] Furthermore, the differentiated control rules are specifically as follows:

[0014] When the stability problem type is thermal stability, if the energy storage power station is in the charging state, it will prioritize reducing the charging power or stopping charging; if it is in the discharging state, it will prioritize reducing the discharging power or stopping discharging.

[0015] When the stability problem type is transient voltage stability, the energy storage station will prioritize reducing power and reducing its reactive power absorption when it is in charging state, and will prioritize quickly switching to reactive power priority mode to increase reactive power generation when it is in discharging state.

[0016] When the stability problem type is transient power angle stability or dynamic stability, priority should be given to adjusting the active or reactive power of the energy storage power station to provide dynamic support or suppress oscillations.

[0017] Furthermore, it is determined whether the regulation capability of the energy storage power station meets the grid stability requirements. Specifically, the minimum control quantity required after a grid fault is calculated. If the adjustable power range of the energy storage power station is less than the minimum control quantity required, it is determined that the regulation capability is insufficient. The power regulation target value of the priority control command is dynamically set according to the correlation between the current energy storage power and the fault overload, and is not lower than zero.

[0018] Furthermore, the multi-dimensional evaluation indicators include:

[0019] Resource type coefficients are set based on the control value and adjustment cost of various grid-connected units;

[0020] The response speed coefficient is adjusted based on the quantitative determination of the command response time of various grid-connected units;

[0021] The electrical position coefficient is determined based on the quantitative effect of the grid-connected unit on mitigating the current stability problem.

[0022] Furthermore, when generating the resection sequence, weights are dynamically assigned to the multi-dimensional evaluation indicators, with a total weight of 1, and the weight ratio of each dimension is adjusted according to the type of stable problem and the degree of urgency; a comprehensive priority score is calculated by combining the influence coefficient of each dimension with the dynamic weights, and the resection sequence is generated by sorting the scores.

[0023] Furthermore, the specific method for the security verification includes:

[0024] The power transfer distribution factor method is used to calculate the power flow impact of the proposed resource removal on critical lines or sections, and to predict whether new overloads will occur.

[0025] Based on voltage-reactive power sensitivity analysis, estimate the change in critical bus voltage after disconnection and determine whether it is below the voltage safety lower limit.

[0026] If the verification reveals a new power flow violation or a critical bus voltage below the safety lower limit, the cut-off sequence will be adjusted, and the grid-connected unit that caused the problem will be moved to the next position or replaced with another suitable grid-connected unit.

[0027] Furthermore, the triggering condition for responding to a preset severe fault in the power grid is as follows: the relay protection signal belongs to the preset severe fault set and satisfies at least one of the following: thermal stability margin ≤ 0, voltage stability margin ≤ voltage critical action threshold, or transient power angle swing prediction instability; the thermal stability margin is the difference between the thermal stability limit of the component and the real-time power flow, and the voltage stability margin is the difference between the real-time bus voltage and the lower limit of voltage stability.

[0028] Furthermore, after executing the control action, the weight coefficients of the multi-dimensional evaluation indicators and the parameters in the differentiated control rules are optimized offline or online based on the control effect data to achieve adaptive optimization of the strategy.

[0029] And, a safety and stability control system for a large-scale energy storage power station after it is connected to the power grid, comprising:

[0030] The fault response and status identification module is used to respond to preset severe faults in the power grid, determine the type of power grid stability problem, and the real-time operating status of the energy storage power station.

[0031] The control decision module is used to generate priority control commands based on the combination of stability problem type and energy storage operating status;

[0032] The unit disconnection sequence generation module is used to dynamically generate a grid-connected unit disconnection sequence based on multi-dimensional evaluation indicators when the energy storage regulation capacity is insufficient.

[0033] The control execution and safety verification module is used to execute control commands or cut-off sequences, and performs safety verification through power flow transfer analysis and voltage response prediction before execution.

[0034] The adaptive optimization module is used to optimize weight coefficients and control rule parameters based on control effect data.

[0035] Compared to existing technologies, this invention and its preferred solutions are fully adapted to the dual functional attributes and rapid adjustment characteristics of large-scale energy storage power stations. By coupling differentiated control rules with the energy storage operating status and the types of grid stability problems, it maximizes the potential of energy storage to "adjust instead of trip," reduces power generation losses and equipment impact caused by traditional tripping strategies, and significantly optimizes control costs. It breaks through the limitations of fixed priorities in traditional tripping strategies by dynamically allocating weights based on multi-dimensional evaluation indicators to generate a tripping sequence, achieving an optimal balance between control effectiveness and adjustment costs, and improving the accuracy and flexibility of the control strategy. Through a pre-execution safety verification mechanism, it effectively prevents control actions from causing new power flow exceedances or voltage stability problems, enhancing the reliability and safety of grid fault handling. Relying on an adaptive optimization mechanism, it can continuously optimize strategy parameters based on actual control effects, making the control logic more aligned with the dynamic changes in grid operation, further improving the robustness and adaptability of the strategy. Overall, this invention improves the grid safety and stability control system after large-scale energy storage is integrated, enhancing the renewable energy absorption capacity and grid operation economy while ensuring the safe and stable operation of the grid. Attached Figure Description

[0036] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0037] Figure 1 This is a flowchart illustrating the implementation of the method in an embodiment of the present invention.

[0038] Figure 2 The power grid structure diagram is used to simulate and verify the embodiments of the present invention. Detailed Implementation

[0039] In the following, specific embodiments of this application will be described in detail with reference to the accompanying drawings. Based on these detailed descriptions, those skilled in the art will be able to clearly understand and implement this application. Without departing from the principles of this application, features from various embodiments can be combined to obtain new implementations, or certain features from some embodiments can be substituted to obtain other preferred implementations.

[0040] To make the features and advantages of the present invention more apparent and understandable, specific embodiments are described below in conjunction with the accompanying drawings:

[0041] This invention addresses the problem that existing power grid security and stability control strategies do not fully consider the operational characteristics (dual functional attributes, rapid adjustment capabilities, etc.) and control potential of large-scale energy storage power stations. It provides a refined, adaptive, and more economical method for formulating and executing a security-controlled power-off (in this invention, "disconnection of grid-connected units") strategy. The core objectives include:

[0042] 1. Quickly and accurately identify the types of stability problems caused by power grid faults, covering thermal stability, transient power angle stability, transient voltage stability, and dynamic stability;

[0043] 2. Based on the real-time operating status (charging / discharging) of the energy storage power station and the grid stability requirements, intelligently select the control mode of "rapid power adjustment (adjustment instead of switching)" or "disconnection of grid-connected units";

[0044] 3. Establish a precise generator switching priority system that integrates multiple factors such as resource type, adjustment speed, adjustment cost, and electrical location to achieve coordinated and optimized control of differentiated resources such as energy storage, adjustable industrial load, thermal power units, and new energy units;

[0045] 4. Systematically integrate the above strategies and connect them to the existing security control system to enhance the power grid's ability to cope with severe faults such as N-2 and N-3.

[0046] The implementation process of this invention is as follows: Figure 1 As shown, the specific steps include:

[0047] Step 1: Real-time monitoring and fault diagnosis

[0048] The security control system's main station collects key power grid data in real time, including traditional operating parameters such as power flow at key sections, line power, bus voltage, and frequency, as well as core information of each controlled power station (operating status of energy storage power stations (charging P)). ch >0, discharge P dis >0), adjustable power range ΔP ess The system calculates the thermal stability margin and voltage stability margin, matches the preset fault set with the relay protection signal, and classifies the severity. When the preset triggering conditions are met (such as power flow exceeding the thermal stability limit, voltage approaching the instability threshold, or predicted transient power angle instability), the subsequent strategy calculation is initiated.

[0049] This step constructs a panoramic real-time monitoring network covering traditional power grid parameters and the dynamic characteristics of energy storage. For the first time, it incorporates the real-time operating status and dynamic adjustability of energy storage into the core input of safety control decision-making, breaking through the simplistic positioning of energy storage in traditional safety control systems and providing data support for subsequent refined and differentiated control.

[0050] Step 2: Quickly identify stable problem types

[0051] Based on the real-time monitoring data from step 1 and the power flow transfer calculation results after the fault, the main stability problems faced by the system and their severity are quickly identified. The identification follows a hierarchical logic: transient power angle stability is prioritized (most urgent), followed by transient voltage stability, then thermal stability, and finally dynamic stability. The final output is a type vector containing the primary and secondary problems and their urgency.

[0052] This step, serving as the "problem diagnosis stage" in strategy formulation, replaces the traditional "one-size-fits-all" extensive approach. By accurately locating the core contradictions of the fault, it provides a crucial basis for selecting appropriate control methods in the future, and is the core prerequisite for upgrading control strategies from extensive to refined.

[0053] Step 3: Preliminary control decisions based on stability type and energy storage state

[0054] Based on the stability problem type determined in step 2 and the real-time operating status of the energy storage power station, a preliminary control decision logic is constructed to form targeted control recommendations. For example, for thermal stability problems, in the charging state of the energy storage, priority is given to reducing power or stopping charging to reduce the load on the power transmission channel, while in the discharging state, reducing power or stopping power generation can alleviate the pressure on the power receiving channel; for transient voltage stability problems, the energy storage can provide voltage support by quickly adjusting reactive power absorption or increasing reactive power generation.

[0055] This step establishes a coupled decision-making mechanism between stability problem types and energy storage operating states, clarifying the priority control actions for energy storage in different scenarios, fully leveraging its flexible adjustment capabilities, and providing clear rule support for the implementation of adjustment-based switching. The decision output is a recommended strategy, including energy storage adjustment instructions, expected effect assessments, and backup control activation flags: if stability requirements can be met solely through energy storage adjustment, the execution phase begins directly; if adjustment capacity is insufficient or grid-connected units need to be disconnected, the next step is triggered.

[0056] Step 4: Establish a precise machine-switching priority order table based on multi-factor fusion.

[0057] When step 3 determines that backup control needs to be activated, the system first determines the optimization direction based on the stability problem type (thermal stability direction focuses on rapid and accurate power flow reduction, while voltage / power angle stability direction focuses on electrical support effect). Then, it acquires the real-time characteristics (cut-off power, response time, electrical position, etc.) of all controllable resources in the "allowed-to-cut" state. By calculating resource type coefficients, power regulation speed coefficients, and electrical position / sensitivity coefficients, and assigning dynamic adaptive weights, a comprehensive scoring model is constructed. Based on the scoring results, resources are sorted in descending order, and a cutting-off sequence is formed by combining the required minimum control quantity. Simultaneously, power flow exceedance prediction and voltage safety verification ensure that the sequence has no secondary risks. Finally, a cutting-off priority table with clearly defined cutting targets, cutting amounts, and execution order is output.

[0058] This step replaces the traditional fixed and experience-based switching sequence in security control. By combining dynamic weights and multi-dimensional coefficients, it achieves on-demand optimization of the switching strategy and minimizes control costs.

[0059] Step 5: Strategy Implementation and Feedback

[0060] This step is the strategy closed-loop implementation stage, which includes three types of operations: First, implementing adjustment instead of switching, sending power adjustment commands (including target values ​​and rates of change) to the energy storage power station and monitoring the adjustment effect in real time; second, implementing precise switching, sending switching commands in rounds according to the priority order table generated in step 4, and evaluating whether the stable state meets the standard after each round of execution; third, adaptive optimization, the system records the actual effect of each control action, and optimizes and adjusts the weight coefficients and basic priority parameters of the comprehensive scoring model offline or online to continuously improve the strategy adaptability and control accuracy.

[0061] As a preferred implementation of this embodiment, step 1 specifically includes the following steps:

[0062] S1.1 Data Acquisition and Input

[0063] The system collects core data on power grid operation and key information from each controlled power plant, specifically including:

[0064] Basic operating parameters of the power grid: power flow at key sections, line transmission power, bus voltage, and system frequency;

[0065] Controlled power plant information: Operating status of energy storage power stations (charging status is marked as P) ch >0, Discharge status is marked as P dis >0) and dynamically adjustable power range (ΔP) ess (); Real-time output and available capacity of each thermal power unit; Actual output and available capacity of each new energy power station; Real-time load size of each controllable industrial load.

[0066] S1.2 Real-time stability margin calculation (core triggering criterion)

[0067] The system's main station continuously performs online stability margin calculations on preset key components and cross-sections, using two types of indicators and taking the calculation results as the core basis for triggering strategies:

[0068] (1) Calculation of thermal stability margin

[0069] Calculation formula: Thermal stability margin (unit: MW) = Current thermal stability limit of component - Real-time power flow value of component;

[0070] Calculation objects: Focus on lines, key sections and core transformers within the jurisdiction of the main station of the system;

[0071] Triggering rules: When the calculated margin value is lower than the preset warning threshold, the system enters the warning state; when the margin value is ≤0 (i.e., the real-time power flow reaches or exceeds the thermal stability limit) and a relevant fault signal is received simultaneously, the subsequent control strategy calculation process is immediately triggered.

[0072] (2) Calculation of voltage stability margin

[0073] Calculation formula: Voltage stability margin (unit: pu) = Real-time bus voltage value - Lower limit of voltage stability;

[0074] Calculation objects: Focusing on the voltage of key 500kV busbars and 220kV busbars in the power grid;

[0075] Triggering rules: In accordance with the "Guidelines for Evaluation of Voltage Stability of Power Systems", when the real-time bus voltage is lower than the preset action threshold (for example, setting 0.85pu as the warning threshold and 0.80pu as the strategy trigger action threshold), or when the bus voltage is predicted to be unstable based on the locally measured voltage drop rate (dV / dt), the strategy calculation is initiated.

[0076] S1.3 Fault Type and Severity Determination

[0077] Combining the switch trip signal with the stability margin calculation results of S1.2, the fault type and system impact level are quickly determined using the "logic tree discrimination method". The specific process is as follows:

[0078] S1.3.1 Fault Diagnosis Logic Flow

[0079] 1. Relay protection signal reception: Real-time acquisition of tripping action information of key power components such as lines and main transformers to identify the initial component where the fault occurred;

[0080] 2. Rapid Calculation of Power Flow Redistribution: Based on the real-time power flow distribution and grid topology before the fault, the distribution factor method or DC power flow approximation method is used to quickly estimate the power flow redistribution of remaining components after the fault. The core calculation formula is:

[0081] ΔP m =GSDF m,n × P loss ; where ΔP m GSDF represents the power flow variation of line m. m,n P represents the power generation transfer distribution factor of line m to faulty line n. loss This represents the transmission power lost after the faulty line goes out of service;

[0082] 3. Preset fault set matching: The "fault component combination mode" is precisely matched with the "calculated most severe power flow limit exceeding object" to clarify the specific type of fault (such as line N-2 fault, main transformer tripping fault, etc.).

[0083] 4. Severity Classification: Based on the degree of exceedance of estimated or measured parameters (including power flow overload percentage and bus voltage drop depth), the faults are classified into different emergency levels. The specific classification criteria are as follows:

[0084] Level 1 (Emergency): Multiple lines experience N-2 or N-3 faults, causing the power flow at critical sections to exceed the short-term allowable thermal stability limit of 105%, or the bus voltage to drop rapidly to below 0.75 pu;

[0085] Level 2 (Severe): A single-circuit N-1 fault or an N-2 fault in a specific scenario causes the power flow to exceed the limit but does not exceed the short-term overload capacity of the equipment, or the bus voltage is on the verge of instability.

[0086] S1.4 Comprehensive Trigger Logic Determination

[0087] The instruction for "triggering control strategy calculation" is generated by the following logical combination to ensure the accuracy of the triggering conditions:

[0088] Trigger signal = (protection trip signal ∈ preset severe fault set) AND ([thermal stability margin ≤ 0] OR [voltage stability margin ≤ 0]) OR [Transient work angle swing prediction instability]), where, This is the critical voltage action threshold, consistent with the judgment criteria for voltage stability margin calculation in S1.2.

[0089] Among them, the prediction of power angle instability can be based on simplified simulation or results from other stability early warning modules.

[0090] S1.5 Data Encapsulation and Transmission

[0091] Upon triggering the policy calculation, the system automatically encapsulates core information into a standard data packet, which is then transmitted to the policy generation module in real time. The data packet contains the following key information:

[0092] Fault identifier: Fault ID (Example format: FAULT_2025_BK_N2);

[0093] Fault Details: Fault type (Example: Baokun Line 3 N-2 fault), key overload objects and specific values ​​(Example: Baokun Line 3, estimated power flow 2532MW, overload 232MW);

[0094] System operation background: System operation mode label (Example: Kunwu Line 1 power outage mode, energy storage fully charged state);

[0095] Control resource base: Real-time severable capacity and current operating status of each execution station (energy storage power station, thermal power unit, controllable industrial load, new energy power station).

[0096] Step 2 specifically includes the following steps:

[0097] S2.1 Thermal Stability Problem Identification: Based on the post-fault steady-state power flow distribution and component thermal capacity constraints, determine whether there is continuous equipment overload due to power transfer. The specific identification process is as follows:

[0098] 1. Obtain power flow distribution after the fault: Based on the real-time data in step 1, quickly perform DC power flow calculation or use the pre-calculated distribution factor to estimate the power flow of the entire network after the fault.

[0099] The line breakage distribution factor is used for rapid calculation:

[0100]

[0101] in For the line after the fault trend, The current flow before the failure. Faulty line The initial trend, For the line For the line The distribution factor of the break.

[0102] 2. Heat capacity verification

[0103] Verify each piece of equipment (lines, transformers) in the power grid that may exceed limits:

[0104] Overload indicator =

[0105] in, For line / transformer The thermal stability limit power after ambient temperature correction.

[0106] 3. Criteria for judging thermal stability issues:

[0107] The main criterion is that if the estimated power flow of any 500kV line exceeds its thermal stability limit after the ambient temperature correction, it is determined to be a thermal stability problem.

[0108] Auxiliary criterion: Also consider the predicted overload duration. If the overload is greater than 5% and the predicted duration is greater than 10 minutes, the problem is more serious.

[0109] S2.2 Transient Power Angle Stability Problem Judgment Method: Based on the numerical solution of the generator rotor motion equation after a fault or the energy function method, determine whether the system can maintain synchronous operation. The specific judgment process is as follows:

[0110] S2.2.1 Simplified Time-Domain Simulation Implementation Description: The system adopts a simplified time-domain simulation model based on nodal admittance matrices and differential-algebraic equations (DAE). The specific process is as follows:

[0111] (1) Model simplification principle: retain the detailed model of the key generator (such as the classical second-order model or the fourth-order detailed model); use the constant impedance model or the static ZIP model for the load; use the quasi-steady-state algebraic equation to describe the network and ignore the electromagnetic transient process.

[0112] (2) Simulation equations:

[0113] Generator rotor motion equation:

[0114] in, Let t be the rotor power angle of generator i, and t be time. Let be the instantaneous angular velocity of the rotor of generator i. For synchronous angular velocity, Let i be the rotor moment of inertia. Let i be the mechanical input power of generator i. Let i be the electromagnetic output power of generator i. Let be the damping coefficient of generator i.

[0115] Network algebraic equations (node ​​power balance):

[0116] in, Let be the total active power output of the generators at node i. Let i be the total active power consumption of the load. and Let be the voltage magnitudes at nodes i and j, respectively. The electrical conductance between node i and node j Let be the voltage phase angle difference between node i and node j. Let n be the electrical susceptance between node i and node j, and n be the total number of nodes in the power grid.

[0117] (3) Simulation process:

[0118] 1. Initialization configuration:

[0119] Importing basic parameters: Loading power flow data, network topology, and unit parameters (moment of inertia M) before the fault. i Damping coefficient D i etc.), load model parameters (constant impedance / static ZIP model coefficients);

[0120] Initial value calculation: Based on the steady-state power flow before the fault, solve for the generator rotor initial power angle, synchronous angular velocity, and electromagnetic power, initialize the load power, node voltage amplitude, and phase angle, and ensure that the node power balance equation is satisfied.

[0121] 2. Simulation during the fault period (0~t) clear , t clear (Time for fault clearing)

[0122] Admittance matrix modification: Adjust the node admittance matrix according to the fault type (such as line N-2 trip, bus short circuit): for trip faults, set the branch admittance of the corresponding element to zero; for short circuit faults, add the short circuit admittance to the fault node.

[0123] State variable solution: Based on the modified admittance matrix, continuously solve the differential-algebraic equations (DAE) system and record the dynamic changes of unit power angle, angular velocity and node voltage during the fault.

[0124] 3. Simulation after fault clearing (t) clear ~t clear +3~5 seconds):

[0125] Network structure restoration: Restore the admittance matrix to its pre-fault state (or adjust it according to the actual topology after the fault is cleared; if a permanent trip occurs, keep the admittance of the faulty element at zero).

[0126] Integral solution: Use the prediction-correction method or semi-implicit integration method (to ensure numerical stability), set the time step (e.g., 1~10ms, balancing accuracy and efficiency), and perform integral solution on the generator rotor motion equation and nodal power balance equation for 3~5 seconds.

[0127] Process constraints: Synchronously monitor changes in unit angular velocity and power angle difference to avoid numerical divergence.

[0128] 4. Results Output and Organization:

[0129] Key unit selection: Select the main thermal power units of the system, the units associated with the new energy collection station, and the units with the highest ranking of the oscillation mode participation factor as key units;

[0130] Power angle difference curve generation: Outputs the relative power angle difference between any two key units. The curve changes over time, with the time axis covering 0.5 seconds before the fault to 3-5 seconds after the fault is cleared, and the data sampling interval is consistent with the integration time step.

[0131] in, Let be the relative power angle difference between unit i and unit j at time t. Let be the rotor power angle of generator i at time t. Let be the rotor power angle of generator j at time t.

[0132] Auxiliary data output: Synchronously output node voltage trajectories and unit electromagnetic power change curves to provide support for stability criterion verification.

[0133] (4) Accelerated calculation methods: The numerical stability can be improved by using the prediction-correction method or the semi-implicit integration method; multiple fault scenarios can be calculated in parallel; and initial state correction based on PMU data is supported.

[0134] (5) Stability criterion for the angle of work:

[0135] The initial swing stability criterion is adopted: if the maximum value of the relative power angle difference between any two machines does not exceed a threshold during the simulation period. If the power angle difference converges at the end of the simulation, then the transient power angle is considered stable.

[0136] Criterion for instability of multiple pendulums: If the first pendulum swing passes through but the subsequent swing amplitude increases, resulting in oscillations with increasing amplitude, then it is judged as dynamic instability.

[0137] S2.3 Transient Voltage Stability Problem Identification Method: Based on the load node voltage recovery capability and reactive power balance after a fault, determine whether voltage collapse will occur. The specific identification process is as follows:

[0138] 1. Voltage trajectory analysis method: Monitor the voltage response curve V(t) of the critical load bus.

[0139] Criterion 1 (Transient voltage drop): If the fault is cleared (corresponding to fault clearing time t) clear After that, the bus voltage failed to recover to above 0.8 pu within the first second: .

[0140] Criterion 2 (Medium- to Long-Term Voltage Recovery): If the voltage remains below 0.90 pu and shows a downward trend within 3-10 seconds after the fault is cleared: .

[0141] 2. Reactive power reserve and sensitivity analysis:

[0142] Calculate the reactive power reserve margin in the critical area: ,

[0143] Among them, Q G,max Q is the sum of the maximum reactive power output of all generators in the region. G,current It is the sum of the current reactive power output of all generators in the region.

[0144] Calculate the voltage-reactive power sensitivity matrix of the load nodes to identify weak nodes.

[0145] Criterion: If the reactive power reserve in a certain area is less than 10%, and the sensitivity of the voltage at critical nodes to reactive power injection is |dV / dQ| > 0.1 pu / MVar, then the system voltage stability margin is insufficient.

[0146] S2.4, Method for determining dynamic stability / small disturbance stability: Based on eigenvalue analysis of the system's linearized model or measured oscillation mode identification, determine whether the system damping is sufficient. The specific determination process is as follows:

[0147] 1. Oscillation pattern recognition based on PMU data:

[0148] The generator power angle, frequency, and line power oscillation data provided by the Wide Area Measurement System (WAMS) are utilized.

[0149] The frequency f and damping ratio ζ of the dominant oscillation mode are extracted using the Prony algorithm or matrix beam method.

[0150] ,

[0151] Where x(t) is the dynamic response signal of the power system being analyzed, M is the number of dominant oscillation modes obtained from the fitting, and A i Let be the amplitude of the i-th dominant oscillation mode. The attenuation factor for the i-th dominant oscillation mode. Let be the initial phase angle of the i-th dominant oscillation mode. Let be the initial phase angle of the i-th dominant oscillation mode.

[0152] 2. Damping ratio criterion:

[0153] Normal requirements: The damping ratio of regional oscillation modes and oscillation modes strongly correlated with major power plants and generating units should reach 0.03 or higher.

[0154] Special measures after a fault: The damping ratio should be at least 0.01~0.015.

[0155] Instability criterion: If any dominant oscillation mode identified satisfies:

[0156]

[0157] This is then classified as a dynamically stable problem.

[0158] Pattern Participation Factor Analysis:

[0159] Calculate the participation factor of each generator in the weakly damped mode and identify the group of units that contribute the most to the oscillation.

[0160] 1. Classification of operating modes:

[0161] Normal mode: The power grid is fault-free, key components (lines / main transformers) are in operation, and the output of new energy sources and load are in a steady-state operation within the normal fluctuation range;

[0162] Special circumstances after a fault: When a fault such as N-1 or N-2 occurs in the power grid, the fault has been cleared but the system is still in a dynamic transition phase (not yet restored to steady state), or a critical component is permanently out of service, resulting in a state of power grid topology adjustment.

[0163] 2. Pattern Classification:

[0164] Regional oscillation mode: Low-frequency oscillation mode (typically 0.1~1.0Hz) between different regional generating units in the power grid, with a wide range of impact;

[0165] Strongly correlated oscillation mode: an oscillation mode that is strongly coupled with the electromechanical characteristics of the main power plants and large-capacity units (such as megawatt-class thermal power units and large-scale new energy collection stations) in the system. Its stability directly determines the dynamic security of the entire network.

[0166] Weakly damped mode: Damping ratio ζ i Oscillation modes below the minimum allowable threshold under the corresponding operating mode (are the core cause of dynamic stability problems).

[0167] 3. Damping ratio requirements for different operating modes

[0168] Normal mode: Regional oscillation mode and oscillation mode strongly correlated with major power plants and generating units, damping ratio ζ i It needs to be ≥0.03 (i.e., 3%).

[0169] Special approach after a fault: Damping ratio ζ of the aforementioned key oscillation modes i The threshold should be ≥0.01~0.015 (i.e. 1%~1.5%). Lowering the threshold takes into account the transient characteristics of the system after a fault, and balances stability and power supply continuity.

[0170] If any dominant oscillation mode identified by the Prony algorithm or matrix bundle method satisfies the following formula:

[0171]

[0172] If the system has dynamic stability issues, targeted damping control needs to be initiated (such as injecting reverse damping power from energy storage and precisely cutting off units that contribute significantly to oscillations).

[0173] In this invention, the mode participation factor is an index (range 0~1) that quantifies the contribution of each generator to a certain weakly damped mode. The core of the calculation is to determine the participation of each unit in the weakly damped oscillation through the eigenvector analysis of the generator rotor motion equation: the closer the participation factor is to 1, the greater the contribution of the unit to the oscillation mode, and the core source of the oscillation.

[0174] By calculating the participation factor of each generator in the weakly damped mode, the generator unit with the largest contribution to oscillation is accurately identified, providing a basis for the priority ranking of generator tripping in the subsequent step 4: For generator units with large participation factors in the weakly damped mode, higher weight is given in the generator tripping ranking of dynamic stability problems (prioritizing tripping or regulation), which can quickly suppress oscillation and minimize control costs.

[0175] S2.5 Comprehensive Judgment Logic and Priority:

[0176] In real-world systems, multiple stability problems may coexist. This invention employs a hierarchical discrimination method:

[0177] First priority: Transient power angle stability — If the simulation shows that the power angle is out of step (δ>180°), it is directly determined to be a transient power angle instability problem, which is the most urgent situation.

[0178] Second priority: Transient voltage stability—If there is no power angle instability, but the critical bus voltage remains below 0.75 pu, it is determined to be a voltage stability problem.

[0179] Third priority: Thermal stability issues—If neither of the first two is present, but there is an overloaded device with an overload value >5%, then it is determined to be a thermal stability issue.

[0180] Fourth priority: Dynamic stability problem — If none of the above apply, but a weakly damped oscillation mode is identified (ζ < threshold), then it is determined to be a dynamic stability problem.

[0181] Step S2 ultimately outputs a stability problem type vector, such as: [Main problem: thermal stability, Secondary problem: voltage stability, Severity: severe] or [Main problem: transient power angle instability, Severity: urgent]. This result will be directly input into step 3 to generate targeted control decisions.

[0182] Step 3 specifically includes the following steps:

[0183] S3.1 Preliminary control decision based on stability type and energy storage status: Based on the stability problem type in step 2 and the real-time status of the energy storage power station, a preliminary control decision matrix is ​​formed.

[0184] This preliminary control decision matrix uses the core scenario dimension of the stable problem type determined in step 2 and the real-time operating status (charging / discharging) of the energy storage power station as the action adaptation dimension. For each combination, it clarifies the corresponding energy storage control actions, specific operating rules, expected control effects, and backup control triggering conditions, as detailed below:

[0185] I. Control Decisions in Thermal Stability Problem Scenarios

[0186] 1. The energy storage is in a charging state (P). ch >0)

[0187] Action type: Adjustment instead of switching (rapid adjustment of active power)

[0188] Specific steps: Immediately reduce the energy storage charging power, and adjust the target value according to formula P. ch,new = max(0, P ch- k×ΔP) calculation (where ΔP is the power that needs to be reduced if the thermal stability limit is exceeded, and k is the safety margin coefficient, which is used to deal with uncertainties such as adjustment error, communication and execution delay, and is preferably 1.2); if the charging power has dropped to 0 and still cannot meet the thermal stability margin requirements, the charging stop operation is executed.

[0189] Expected results: Reduce the active load on the power transmission channels and quickly raise the thermal stability margin above the warning threshold.

[0190] Backup control trigger: If the thermal stability margin is still ≤0 after adjustment / stop charging, the precise machine switching process in step 4 will be triggered.

[0191] 2. The energy storage is in a discharging state (P dis >0)

[0192] Action type: Adjustment instead of switching (rapid adjustment of active power)

[0193] Specific steps: Immediately reduce the energy storage discharge power, and adjust the target value according to formula P. dis,new = max(0, P dis Calculate -k×ΔP); if the discharge power has dropped to 0 and still cannot meet the thermal stability margin requirement, execute the shutdown operation.

[0194] Expected effect: Reduce the active load of the power receiving channel and quickly increase the thermal stability margin to above the warning threshold.

[0195] Backup control trigger: If the thermal stability margin is still ≤0 after adjustment / stop, the precise machine switching process in step 4 will be triggered.

[0196] II. Control Decisions in Transient Power Angle Stability Problem Scenarios

[0197] 1. The energy storage is in a charging state (P). ch >0)

[0198] Action type: Adjustment instead of switching (rapid adjustment of active power)

[0199] Specific steps: Immediately increase the energy storage charging power, and adjust the target value according to formula P. ch,new = min(P ch,max , P_ch + k×ΔP osc ) Calculate (where ΔP) osc The power that needs to be adjusted to suppress power angle oscillation, P ch,new (Maximum charging power for energy storage).

[0200] Expected effect: Increase the system's equivalent damping and suppress the further expansion of the power angle difference of key units.

[0201] Backup control triggered: The power angle difference of the key unit continues to increase after adjustment (δ)ij If (t) > 120°, then the precise machine cutting process in step 4 will be triggered.

[0202] 2. The energy storage is in a discharging state (P dis >0)

[0203] Action type: Adjustment instead of switching (rapid adjustment of active power)

[0204] Specific operation: Immediately increase the energy storage discharge power, and adjust the target value according to formula P. dis,new = min(P dis,max ,P dis + k×ΔP osc Calculate, where P dis,max This represents the maximum discharge power of the energy storage.

[0205] Expected effect: Increase the system's equivalent damping and suppress the further expansion of the power angle difference of key units.

[0206] Backup control triggered: The power angle difference of the key unit continues to increase after adjustment Δ ij If (t) > 120°, then the precise machine cutting process in step 4 will be triggered.

[0207] III. Control Decisions in Transient Voltage Stability Problem Scenarios

[0208] 1. The energy storage is in a charging state (P). ch )

[0209] Action type: Adjustment instead of switching (rapid adjustment of reactive power)

[0210] Specific steps: Immediately reduce the active power charging power of the energy storage, release the reactive power regulation capacity, and simultaneously increase the reactive power output of the energy storage to Q. dis,max (Maximum reactive power output of energy storage).

[0211] Expected effect: Improve the bus voltage in the fault area and prevent the voltage from dropping rapidly below 0.8 pu.

[0212] Backup control trigger: If the bus voltage is still <0.9pu and dV / dt<0 within 5 seconds after adjustment, the precise machine switching process in step 4 will be triggered.

[0213] 2. The energy storage is in a discharging state (P dis >0)

[0214] Action type: Adjustment instead of switching (rapid adjustment of reactive power)

[0215] Specific operation: Immediately reduce the active power discharge of the energy storage, release the reactive power regulation capacity, and simultaneously increase the reactive power output of the energy storage to Q. dis,max .

[0216] Expected effect: Improve the bus voltage in the fault area and prevent the voltage from dropping rapidly below 0.8 pu.

[0217] Backup control trigger: If the bus voltage is still <0.9pu and dV / dt<0 within 5 seconds after adjustment, the precise machine switching process in step 4 will be triggered.

[0218] IV. Control Decisions in Dynamic Stability Problem Scenarios

[0219] 1. The energy storage is in a charging state (P). ch >0)

[0220] Action type: Adjustment instead of cutting (damped power injection)

[0221] Specific operation: Based on the frequency f of the dominant oscillation mode i The damping power that controls the energy storage output is opposite to the oscillation power, and the power amplitude is calculated according to the formula P. damp = K damp ×A osc Calculate (where K) damp The damping control coefficient, whose value is determined through offline simulation tuning or online adaptive algorithm, ensures that the injected damping power can increase the damping ratio of the system's dominant oscillation mode to above the target value. osc (This refers to the oscillation amplitude); synchronous fine-tuning of charging power tracks the oscillation phase.

[0222] Expected effect: Increase the damping ratio of the weakly damped mode to ζ. i ≥ζ min (Normal method ζ) min =0.03, special method after fault ζ min =0.01).

[0223] Backup control trigger: Damping ratio still <ζ after adjustment min If the oscillation amplitude continues to increase, the precise machine switching process in step 4 will be triggered.

[0224] 2. The energy storage is in a discharging state (P dis >0)

[0225] Action type: Adjustment instead of cutting (damped power injection)

[0226] Specific operation: Based on the frequency f of the dominant oscillation mode i The damping power that controls the energy storage output is opposite to the oscillation power, and the power amplitude is calculated according to the formula P. damp = K damp ×A osc Calculate; synchronize fine-tuning of discharge power to track oscillation phase.

[0227] Expected effect: Increase the damping ratio of the weakly damped mode to ζ. i ≥ζ min .

[0228] Backup control trigger: Damping ratio remains ζ after adjustment min If the oscillation amplitude continues to increase, the precise machine switching process in step 4 will be triggered.

[0229] S3.2 Decision Output and Linkage Interface

[0230] This step serves as the connecting point between Step 3 and subsequent control processes. Its purpose is to transform the preliminary control decisions from Step 3 into a structured "recommended strategy package" and clarify the triggering logic for subsequent processes. The specific content and flow rules are as follows:

[0231] (a) The strategy package is a structured collection of information, containing three key types of content:

[0232] 1. Energy storage preferred regulation command

[0233] The specific control actions for energy storage power stations are clearly defined in the format "Object: ESS_1; Command type: Charging power reduction; Target value: XX MW; Execution priority: High", which clearly defines the object of adjustment, operation type, quantitative target and execution priority.

[0234] 2. Pre-assessment of adjustment effect

[0235] Based on a simplified power flow / stability fast calculation model, the degree of improvement of key power grid indicators after the implementation of the regulation command is estimated, such as: "the power flow of overloaded lines is expected to decrease by XX MW, and the voltage of weak bus is expected to rise to XX pu", providing a benchmark for the effect verification of subsequent implementation stages.

[0236] 3. Backup strategy trigger flag

[0237] When it is determined that single energy storage regulation cannot solve the current stability problem (e.g., the overload ΔP is much greater than the adjustable energy storage capacity ΔP), ess If the voltage drop exceeds the reactive power support capacity of the energy storage, then add a trigger flag for "Step 4 Integrated Switching Strategy needs to be started" to the strategy package.

[0238] (ii) Triggering of subsequent processes

[0239] After the strategy package is output, the corresponding subsequent steps are triggered based on its content:

[0240] If the strategy package only contains the "Energy Storage Preferred Regulation Command" (without a backup flag), proceed directly to step 5 and execute the regulation command.

[0241] If the strategy package includes the flag "Step 4 Integrated Switching Strategy Required", then Step 4 is triggered, initiating the multi-resource collaborative switching priority sorting process.

[0242] When step 4 is triggered, the energy storage power station will be included in the cut-off queue as a cut-off unit. However, it is necessary to simultaneously transmit its status information that "adjustment has been performed but the expected effect has not been achieved". This information will be used as the input parameter of the step 4 queue model to reduce the priority weight of energy storage in the cut-off queue (i.e., it will not be prioritized as a cut-off target) in order to maximize the preservation of the resource value that it has participated in adjusting.

[0243] Step 4 specifically includes the following steps:

[0244] S4.1 Backup control activation judgment and mode selection:

[0245] Input: The "Recommended Strategy Package" from step 3, especially the "Backup Strategy Flags" and "Stable Problem Types".

[0246] 1. Triggering conditions: This step is initiated when any of the following conditions are met: (1) Backup strategy flag = TRUE (i.e., energy storage regulation is insufficient to solve the problem). (2) Step 3 directly determines that it is a certain specific serious fault (such as multiple lines N-3), and multi-resource collaborative disconnection needs to be initiated immediately.

[0247] 2. Control Mode Selection: Based on the type of stable problem, set the optimization direction for this round of priority ranking:

[0248] Mode A (Thermal Stability Oriented): The core principle is to reduce the power flow at the overload section as quickly and accurately as possible. The optimization objectives are to minimize the overload and minimize the action delay.

[0249] Mode B (Voltage / Power Angle Stability Guidance): The core is to provide electrical support / damping. The optimization objective is to maximize reactive power support for weak nodes or maximize the negative damping contribution to the dominant oscillation mode. Output: Control Mode = {Mode A or Mode B}.

[0250] S4.2 Acquisition and preprocessing of panoramic information on cuttable resources: Establishing a unified, real-time, and computable feature profile for all "candidate objects" that may be cut:

[0251] Input: Real-time monitoring data from step 1; power grid model and topology; preset unit / load group grouping information. Processing content: (1) Generate candidate list: List all controllable resources that are in the "allowed" state and have the correct power flow direction, including: each energy storage power station sub-unit, controllable industrial load group, thermal power unit, new energy power station / collector line. (2) Extract real-time feature vector: Extract key features for each resource i in the list: Real-time switchable active power (MW); Real-time switchable / adjustable reactive power capacity (MVar); Electrical location (affiliated plant, grid-connected bus). : Estimated total shutdown time (ms, from the time the command is issued until the power reaches zero). Special states (e.g., whether the energy storage is in the "partial adjustment attempted" state).

[0252] Output: List of slicable resource characteristics Resource_List={ ( , , , , ,...)}.

[0253] S4.3 Multi-dimensional influence coefficient calculation: Quantify and score each resource from different dimensions to provide a basic "indicator score" for comprehensive ranking.

[0254] Input: Resource_List; Control mode; Current fault information (overload section, weak bus, dominant oscillation mode).

[0255] Calculation content (each resource is calculated independently):

[0256] 1. Resource type coefficient ( ): Assigning values ​​based on the inherent attributes of resources, reflecting their control value and cost. Recommended baseline assignment range: Conventional industrial load: ≈0.9-1.0. The cost is relatively low, and the impact on the system is small. Energy storage power station unit: ≈0.7-0.9. High regulation value but direct cut-off is costly (power loss, lifespan reduction). If it has already participated in regulation ( This can be achieved by increasing the coefficient to indicate "priority protection". Conventional thermal power units: ≈0.4-0.7. Disconnection involves unit safety, fuel costs, and restart complexity. New energy power plants: ≈0.1-0.4. Cutting off leads to a waste of clean energy and may cause localized voltage problems.

[0257] 2. Power regulation speed coefficient ( ): Scored based on response time, the faster the higher. =exp(- ), where τ is the time constant used for normalization, which can be set according to the typical requirements of the system for control speed (e.g., 100ms). Alternatively, it can be directly normalized. =( ) / ( ).

[0258] 3. Electrical position / sensitivity coefficient ( This is the most critical and dynamic coefficient, measuring the direct effect of removing this resource on resolving the current stability problem, including:

[0259] (1) For mode A (thermal stability): Calculate the absolute value of the power transfer distribution factor (GSDF) of the resource to the overload section. =|GSDF i The larger the value, the more effective it is to alleviate overload by removing the resource.

[0260] (2) For Mode B (voltage / power angle stability): Voltage stability: Calculate the electrical distance or voltage-reactive power sensitivity from the bus where the resource is located to the bus with weak voltage. Power angle stability: The participation factor of the computer group in the dominant oscillation mode. The larger the participation factor, The higher the value.

[0261] S4.4 Dynamic Weight Adaptation and Comprehensive Priority Score Calculation: Based on the urgency and type of the current stable problem, the weights of the coefficients of each dimension are dynamically adjusted, and the comprehensive score of each resource is calculated to achieve precise ranking for "targeted treatment".

[0262] 1. Input: the resources , , Control mode.

[0263] 2. Calculation process:

[0264] Determine the dynamic weights (α, β, γ): the sum of the weights α + β + γ = 1. Their values ​​are determined by the control mode and the severity of the problem.

[0265] (1) Mode A (severe thermal stability): emphasizes "accurate and fast cutting". Set γ (position weight) to be the highest, β (velocity weight) to be the second highest, and α (type weight) to be the lowest. For example: (α,β,γ)=(0.2,0.3,0.5).

[0266] (2) Mode B (insufficient damping): emphasizes "effective cutting (large participation factors)". Set γ (the weight of the participation factors) to be the highest, and α and β to be lower.

[0267] (3) Mode A (slightly thermally stable): may give more consideration to economic efficiency and appropriately increase α (type weight).

[0268] Calculate the overall priority score:

[0269] ,in Normalize to [0,1].

[0270] 3. Output: Overall score for each resource .

[0271] S4.5 Sorting, Verification and Final Machine Switching Sequence Generation: Generate an executable machine switching sequence list based on the score and perform security verification.

[0272] 1. Input: All resources and .

[0273] 2. Processing flow:

[0274] (1) Sort in descending order: Sort the Resources List Press S i Sort by score from highest to lowest. A higher score means a higher "cost-effectiveness" (efficacy / cost) in removing the resource, and it will appear higher in the list.

[0275] (2) Generate cumulative cutting capacity: Calculate the cumulative cuttable capacity ΣP along the sorted list. i .

[0276] (3) Matching stable demand: ΣP i The required minimum control quantity (ΔP) calculated in steps S2 / S3 need (Compare)

[0277] (4) Generate the final sequence: Starting from the top of the sorted list, select the first sequence that satisfies ΣP i ≥ΔP need A subset of resources is used as the first round of machine switching sequence. Multiple backup sequences can be set.

[0278] (5) Safety verification: To assess whether resource removal according to the switching sequence will cause new power flow limit violations or voltage problems, the system adopts a fast verification algorithm based on the power flow transfer distribution factor. The specific steps are as follows:

[0279] a. Construct the resection effect vector:

[0280] For the set of resources R to be removed cut ={R1,R2,...,R m}, calculate its influence vector on the power transfer distribution factor (GSDF) of each critical line / section in the system: Wherein, ΔP lines GSDF is the power flow change vector for all critical paths. line,i Let P be the power generation transfer distribution factor for the i-th cut-off resource of the line. cut,i Let be the power of the removal of the i-th resource.

[0281] b. Trend Limitation Prediction:

[0282] Calculate the power flow P of each path after removal line,new =Pline,current +ΔP line Determine if a new overload has occurred: If P line,new >P limit ×k short If P is the value, then it is considered overloaded. line,new The estimated active power flow value of a certain line after the target resource set is removed; P line,current The current real-time active power flow value of this line before the target resource is removed; ΔP line The change in active power flow of the line after the target resource set is removed; P limit This is the thermal stability limit power of the line after ambient temperature correction; k short The short-term overload factor takes into account the equipment's allowable short-term (e.g., 15-30 minutes) overload capacity, and is preferably 1.05-1.15.

[0283] c. Voltage safety verification (if voltage stability issues are involved):

[0284] Using the voltage-reactive power sensitivity matrix S VQ Estimate the voltage change of the critical bus after disconnection ΔV=S VQ ×ΔQ cut (ΔQ) cut (This refers to the reactive power change caused by the cutoff). Determine if the voltage is below the allowable lower limit (e.g., 0.90 pu).

[0285] d. Verification result processing:

[0286] If the verification passes, the current switching sequence is confirmed; if a new limit violation occurs, the sequence fine-tuning mechanism is activated, prioritizing the shifting of resources that may cause new problems to the later stages, or selecting alternative resource combinations.

[0287] 3. Output: Machine switching priority order table = [(R a , cut P a ),(R b , cut P b This table explicitly specifies the execution station name, the amount of data to be removed, and the execution order for the objects to be removed.

[0288] Step 5: Strategy Execution, Feedback, and Closed-Loop Optimization

[0289] This step is the core of strategy implementation, encompassing two types of control execution: adjustment-based switching and precise machine switching; real-time effect verification; and adaptive strategy optimization, forming a complete closed loop. The specific process is as follows:

[0290] S5.1 Control Execution

[0291] 1. Mode 1: Execute "adjustment instead of switching" (corresponding to the adjustment command in step 3 without a backup flag)

[0292] Command issuance: The master station generates standardized adjustment commands for the energy storage power station, specifying core parameters: power target value (e.g., charging power reduced to 1233MW), adjustment change rate (e.g., maximum change rate ≥100MW / 100ms), execution timeout threshold (e.g., 300ms), and check code (to ensure command transmission security), and issues them to the energy storage execution station through a dedicated security control channel.

[0293] Real-time monitoring: The main station continuously collects energy storage power response data (such as whether the power drops to 1300MW at 250ms and whether it drops to 1250MW at 300ms), and simultaneously monitors the changing trends of key grid indicators (overloaded line power flow, weak bus voltage, system frequency).

[0294] Effect assessment:

[0295] Target achievement: If the energy storage power is stable within ±5% of the target value within the timeout threshold, and the key grid indicators are restored to the safety threshold (such as line power flow < thermal stability limit, voltage > 0.95pu), the adjustment is considered successful, and the system enters the stable monitoring state.

[0296] If the target value is not reached within the time limit, or if the power grid indicators do not improve, the backup generator tripping strategy in step 4 will be triggered immediately to prevent the stability problem from escalating.

[0297] 2. Mode 2: Execute "Precise Machine Switching" (corresponding to the machine switching priority table generated in step 4)

[0298] Instructions are issued in a hierarchical manner: The master station sends the cut-off instructions to the execution stations (thermal power, new energy, controllable load, etc.) in rounds according to the priority order table. Each round of instructions specifies: the cut-off target (e.g., a thermal power unit), the cut-off power (e.g., 50MW), and the execution time limit (e.g., 200ms).

[0299] Inter-round assessment: After each round of shearing is completed, stability verification is performed at intervals of 100~200ms. Through power flow transfer calculation and stability margin review, it is determined whether the stability requirements are met (e.g., thermal stability margin ≥ 0, power angle difference < 120°).

[0300] Target achieved: Subsequent rounds of cutoffs cease, and the system transitions to stable monitoring.

[0301] If the target is not met: Continue to issue the next round of cut-off instructions until the stability requirements are met or the cut-off sequence is exhausted (at this time, the emergency backup control plan will be activated).

[0302] Post-disconnection monitoring: Continuously monitor the status of the disconnected unit (such as whether it is completely disconnected), and whether the power grid experiences new power flow exceeding limits or voltage drops (to avoid secondary faults).

[0303] S5.2 Effect Recording and Adaptive Optimization

[0304] Data archiving: The system automatically records all data for each control action, including: fault type, control command parameters (target value, rate of change, cut-off amount), response time, improvement of grid indicators (such as power flow reduction, voltage recovery value), and control costs (such as energy storage life loss, power generation loss).

[0305] Strategy optimization:

[0306] Online optimization: For the same type of fault (e.g., N-2 line tripping), if the control effect deviation exceeds 10% for three consecutive times (e.g., the power flow still exceeds 5% after adjustment), the weight coefficients (α, β, γ) in step 4 are fine-tuned in real time (e.g., in a thermal stability scenario, the position weight γ is adjusted from 0.5 to 0.55). As a preferred implementation method, the initial values ​​of the dynamic weights (α, β, γ) can be determined based on historical operating data, typical fault simulation results, or expert experience. The core principle is: in thermal stability problems, the focus is on the direct effect of control measures on the overload section (γ is higher) and the execution speed (β is secondary); in voltage / power angle stability problems, the focus is on the support effect of control measures on weak electrical nodes or the ability to suppress the dominant oscillation mode (γ is higher). The weight values ​​can be tuned through offline simulation before the system is put into operation and serve as the basis for subsequent online optimization.

[0307] Offline optimization: Regularly (e.g., monthly) summarize all control cases, optimize basic priorities (e.g., adjust the value range of energy storage type coefficient K_type,i) and stability criterion thresholds (e.g., voltage action threshold) based on big data analysis, update the strategy library, and improve the accuracy and economy of subsequent control.

[0308] For the above online and offline optimization, rule-based self-adjustment methods, gradient descent, reinforcement learning algorithms, or other machine learning methods can be adopted. Historical control effect data (such as the deviation between the actual overload reduction and the target reduction, control costs, etc.) can be used as feedback to iteratively optimize parameters such as weight coefficients (α, β, γ) and resource type coefficients, so that the comprehensive scoring model continuously approaches the actual optimal control effect.

[0309] S5.3 Stability Monitoring and Closed-Loop Closure

[0310] After the control execution meets the target, the system enters a 5-10 second stable monitoring period to continuously monitor the power flow, voltage, frequency and unit power angle difference of the power grid, and confirm that there is no rebound or new stability risk.

[0311] After the monitoring period ends, a complete control report is generated, including fault information, control process, execution data, effect evaluation, and optimization suggestions, providing a basis for subsequent strategy iteration and project review.

[0312] This invention addresses the safety and stability control and tripping strategy problem after a large-scale energy storage power station is connected to the power grid. To verify the correctness of the above control strategy, a 1.5 million kW independent energy storage power station was put into operation at a 500 kV substation in a provincial power grid (power grid structure diagram shown). Figure 2 As shown in the figure, the safety and stability control system and strategy setpoint adaptability of one of the 500kV longitudinal channels of the power grid were simulated and verified:

[0313] 1. Initial running status:

[0314] System operation mode: KW line planned power outage for maintenance, BT area new energy large-scale generation (high output of wind power and photovoltaic).

[0315] Energy storage power station status: The KDL 1.5 million kW independent energy storage power station (energy storage collection station) is in a fully charged state, charging at a power of approximately 1513MW through the CNDZ~KDL 500kV line.

[0316] Power flow at key sections: Initial power of BK three-circuit line: 2841MW; HB section: 4948MW.

[0317] 2. Fault events:

[0318] Fault type: N-2 fault occurred on the BK three-circuit line (two of the three circuits tripped simultaneously).

[0319] Failure time: Assume it is 14:30 on a certain day, during a period of high system load operation.

[0320] Protection action: The circuit breakers at both ends of the faulty line trip within 0.1 seconds.

[0321] 3. Consequences of the failure (traditional analysis)

[0322] Power Flow Transfer: Before the fault, all 2841MW of power was transferred to the remaining BK line.

[0323] Thermal stability issue: The remaining power flow of the BK line reaches 2532MW, exceeding the thermal stability limit of 2300MW under an ambient temperature of 40 degrees Celsius for 500kV lines, with an overload of 232MW (overload rate of 10.1%).

[0324] System risk: If no measures are taken, the line may experience overheating and increased sag due to continuous overload, eventually leading to a chain reaction of trips and expanding the scope of the accident.

[0325] 4. Implementation process of the security control strategy based on the present invention

[0326] 4.1 Step S1: Real-time monitoring and fault diagnosis (0-100ms)

[0327] (1) Real-time monitoring data: WJ master station received the "BK I, II line protection trip" signal (N-2 fault confirmation).

[0328] (2) Real-time power flow monitoring shows that the power of another BK line increased sharply from 947MW to 2532MW (system estimate).

[0329] KDL energy storage power station charging power: 1513MW (P_ch>0).

[0330] KDL bus voltage: 510kV (normal range).

[0331] (3) Stability margin calculation:

[0332] Thermal stability margin calculation: Margin = 2300 - 2532 = -232MW (severe over-limit);

[0333] Voltage stability margin: The voltage of key busbars is all >0.95pu, and there is no risk to voltage stability.

[0334] (4) Triggering strategy calculation:

[0335] Fault matching: Matched the preset scenario "BK three-circuit N-2 fault under KW line outage mode".

[0336] Triggering conditions are met: Protection trip signal ∈ severe fault set AND thermal stability margin ≤ 0 → trigger strategy calculation.

[0337] 4.2 Step S2: Quick identification of stable problem types (100-150ms)

[0338] (1) Determining thermal stability:

[0339] Distribution factor rapid estimation: Calculated using the pre-stored LODF matrix: LODF(BK residual line, fault double loop)≈0.89; Verification: 2532≈947+0.89×(2841-947)=2530MW (consistent with actual measurement).

[0340] (2) Overload assessment:

[0341] Overload rate: 232 / 2300 × 100% = 10.1%;

[0342] Estimated duration: If left uncontrolled, the overload will continue until the current adjusts naturally (several minutes).

[0343] Other stability issues eliminated: Power angle stability: relative angle difference of key units <30°, initial swing stability;

[0344] Voltage stability: Voltage of all 500kV busbars >0.95pu;

[0345] Dynamically stable: No signs of low-frequency oscillation.

[0346] 4.3 Step S3: Preliminary control decision based on stability type and energy storage state (150-200ms)

[0347] (1) Decision matrix query:

[0348] Input: Problem type = thermal stability, energy storage state = charging (P_ch>0);

[0349] Matrix output: Priority control suggestion → "Activate energy storage to reduce power / stop charging".

[0350] (2) Decision-making logic refinement:

[0351] Control target calculation: Power reduction required: ΔP need =232MW (minimum theoretical value);

[0352] Considering safety margin: ΔP target =232×1.2≈280MW (20% margin);

[0353] Energy storage regulation feasibility assessment: Current energy storage charging power: 1513MW;

[0354] Adjustable range: 0~1513MW (full capacity coverage);

[0355] Adjustment speed: ≥100MW / 100ms (meets the speed requirement);

[0356] 4.4 Step S4: Establish a precise priority order table for switching units based on multi-factor fusion (for backup, not triggered in this example) Note: Since step S3 determines that the problem can be solved by energy storage regulation alone, this step is not triggered.

[0357] 4.5 Step S5: Strategy Execution and Feedback (200-500ms)

[0358] 1. Command issuance (200-220ms):

[0359] The WJ master station sends the following command frame to the KDL energy storage power station execution station via the dedicated security control channel: Command frame = {"Command":"Power Adjustment","Target Value":1233, / / MW"Rate of Change":"Maximum","Timeout":300ms,"Check Code":0xA3B5}.

[0360] 2. Process monitoring (220-350ms):

[0361] 250ms: Actual energy storage power drops to 1300MW;

[0362] 300ms: Reduced to 1250MW;

[0363] 320ms: Stable within the range of 1235±5MW.

[0364] 3. Effect verification (350-400ms):

[0365] BK line power flow monitoring values:

[0366] Initial capacity after the fault: 2532MW;

[0367] At 300ms: 2260MW;

[0368] At 400ms: 2248MW (<2300MW, safe);

[0369] System voltage: All buses > 0.95 pu;

[0370] Frequency: 50.05Hz (normal).

[0371] 4. Control success confirmation (400-500ms):

[0372] The main station received a confirmation signal that the energy storage system had completed its adjustment.

[0373] The system entered a stable monitoring state and continued monitoring for 5 seconds without any abnormalities.

[0374] Generate a control report.

[0375] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

[0376] This invention is not limited to the preferred embodiment described above. Anyone inspired by this invention can derive other methods and systems for the safe and stable control of large-scale energy storage power stations connected to the power grid. All equivalent variations and modifications made within the scope of the claims of this invention shall fall within the scope of this invention.

Claims

1. A method for safe and stable control of a large-scale energy storage power station after it is connected to the power grid, characterized in that, include: In response to pre-set severe faults in the power grid, the type of stability problem faced by the power grid is determined, and the real-time operating status of large-scale energy storage power stations connected to the power grid, including charging or discharging status, is obtained. Based on preset differentiated control rules, and combined with the combination of the stability problem type and the real-time operating status of the energy storage power station, priority control instructions are generated. These priority control instructions are used to schedule the energy storage power station to perform rapid power adjustment. If the regulation capacity of the energy storage power station is insufficient to meet the grid stability requirements, a grid-connected unit disconnection sequence is dynamically generated based on a multi-dimensional evaluation index consisting of resource type, regulation response speed and electrical location. Control actions are executed according to the aforementioned priority control commands or cut-off sequences. Before execution, safety checks are performed through power flow transfer analysis and voltage response prediction to avoid triggering additional power grid stability problems.

2. The safety and stability control method for large-scale energy storage power stations after grid connection according to claim 1, characterized in that: The stability problem types include transient power angle stability, transient voltage stability, thermal stability, and dynamic stability. A hierarchical discrimination logic is used to determine the priority, with the specific priority order being transient power angle stability > transient voltage stability > thermal stability > dynamic stability.

3. The safety and stability control method for large-scale energy storage power stations after grid connection according to claim 1, characterized in that: The specific differential control rules are as follows: When the stability problem type is thermal stability, if the energy storage power station is in the charging state, it will prioritize reducing the charging power or stopping charging; if it is in the discharging state, it will prioritize reducing the discharging power or stopping discharging. When the stability problem type is transient voltage stability, the energy storage station will prioritize reducing power and reducing its reactive power absorption when it is in charging state, and will prioritize quickly switching to reactive power priority mode to increase reactive power generation when it is in discharging state. When the stability problem type is transient power angle stability or dynamic stability, priority should be given to adjusting the active or reactive power of the energy storage power station to provide dynamic support or suppress oscillations.

4. The safety and stability control method for large-scale energy storage power stations after grid connection according to claim 1, characterized in that: To determine whether the regulation capability of the energy storage power station meets the grid stability requirements, specifically, the minimum control quantity required after a grid fault is calculated. If the adjustable power range of the energy storage power station is less than the minimum control quantity required, it is determined that the regulation capability is insufficient. The power regulation target value of the priority control command is dynamically set according to the correlation between the current energy storage power and the fault overload, and is not lower than zero.

5. The safety and stability control method for large-scale energy storage power stations after grid connection according to claim 1, characterized in that: The multi-dimensional evaluation indicators include: Resource type coefficients are set based on the control value and adjustment cost of various grid-connected units; The response speed coefficient is adjusted based on the quantitative determination of the command response time of various grid-connected units; The electrical position coefficient is determined based on the quantitative effect of the grid-connected unit on mitigating the current stability problem.

6. The safety and stability control method for a large-scale energy storage power station after grid connection according to claim 5, characterized in that: When generating the resection sequence, weights are dynamically assigned to the multi-dimensional evaluation indicators, with a total weight of 1. The weight ratio of each dimension is adjusted according to the type of stable problem and the degree of urgency. A comprehensive priority score is calculated by combining the influence coefficient of each dimension with the dynamic weights, and the resection sequence is generated by sorting the scores.

7. The safety and stability control method for large-scale energy storage power stations after grid connection according to claim 1, characterized in that: The specific methods for security verification include: The power transfer distribution factor method is used to calculate the power flow impact of the proposed resource removal on critical lines or sections, and to predict whether new overloads will occur. Based on voltage-reactive power sensitivity analysis, estimate the change in critical bus voltage after disconnection and determine whether it is below the voltage safety lower limit. If the verification reveals a new power flow violation or a critical bus voltage below the safety lower limit, the cut-off sequence will be adjusted, and the grid-connected unit that caused the problem will be moved to the next position or replaced with another suitable grid-connected unit.

8. The safety and stability control method for a large-scale energy storage power station after grid connection according to claim 1, characterized in that: The triggering conditions for responding to the preset severe faults in the power grid are as follows: the relay protection signal belongs to the preset severe fault set and satisfies at least one of the following: thermal stability margin ≤ 0, voltage stability margin ≤ voltage critical action threshold, or transient power angle swing prediction instability; the thermal stability margin is the difference between the thermal stability limit of the component and the real-time power flow, and the voltage stability margin is the difference between the real-time bus voltage and the lower limit of voltage stability.

9. The safety and stability control method for a large-scale energy storage power station after grid connection according to claim 1, characterized in that: After executing the control action, the weight coefficients of the multi-dimensional evaluation indicators and the parameters in the differentiated control rules are optimized offline or online based on the control effect data to achieve adaptive optimization of the strategy.

10. A safety and stability control system for a large-scale energy storage power station after it is connected to the power grid, characterized in that, include: The fault response and status identification module is used to respond to preset severe faults in the power grid, determine the type of power grid stability problem, and the real-time operating status of the energy storage power station. The control decision module is used to generate priority control commands based on the combination of stability problem type and energy storage operating status; The unit disconnection sequence generation module is used to dynamically generate a grid-connected unit disconnection sequence based on multi-dimensional evaluation indicators when the energy storage regulation capacity is insufficient. The control execution and safety verification module is used to execute control commands or cut-off sequences, and performs safety verification through power flow transfer analysis and voltage response prediction before execution. The adaptive optimization module is used to optimize weight coefficients and control rule parameters based on control effect data.

Images