A power grid stability control method based on real-time risk assessment

Abstract

The application discloses a power grid stability control method based on real-time risk assessment, relates to the field of power system automation and intelligent control, and aims to solve the problems of traditional stability control, such as dependence on static threshold, response lag and lack of foresight. The method comprises the following steps: collecting multi-source operation data in real time to form a unified time reference multi-dimensional state vector; constructing a dynamic risk assessment model containing voltage deviation, frequency change rate and fault current characteristics, and calculating a comprehensive risk index; identifying risk hotspots and propagation paths in combination with power grid topology; generating control instructions such as generator tripping, load shedding or reactive power compensation according to the risk level; and iteratively optimizing the control strategy through closed-loop feedback until the system recovers to stability. Through the technical scheme, the application realizes risk-driven accurate, fast and adaptive stability control, and significantly improves the safety margin, response speed and economy of a complex power grid under disturbance.

Description

Technical Field

[0001] This invention relates to the field of power system automation and intelligent control, and in particular to a power grid stability control method based on real-time risk assessment. Background Technology

[0002] With the continuous expansion of smart grid scale and the increasing complexity of its operation, grid stability control has become a core component in ensuring the safe and reliable operation of the power system. Modern power grids, influenced by factors such as the high proportion of renewable energy integration, intensified load fluctuations, and multi-regional interconnections, exhibit highly nonlinear and strongly coupled dynamic behavior, posing unprecedented challenges to the real-time performance, accuracy, and adaptability of control strategies. Especially in the initial stages of faults or disturbances, the system state evolves rapidly, necessitating a technological means to quickly identify potential instability risks and dynamically adjust control measures.

[0003] However, existing power grid stability control methods largely rely on offline simulation or preset threshold triggering mechanisms, making it difficult to conduct continuous and quantitative risk assessments of operating conditions. Traditional control strategies are typically based on static safety margin indicators, failing to effectively integrate multi-source real-time measurement data (such as PMU phasors, switch states, and frequency deviations) to construct dynamic risk models. Furthermore, existing methods lack quantitative representations of risk propagation paths and key weaknesses, leading to lagging or overly conservative control commands. In addition, the control decision-making process often ignores the cumulative effect of risk evolution over time, failing to establish a mathematical mapping relationship between risk levels and control action intensity, making the system prone to undercontrol or increased oscillations in critical states. Summary of the Invention

[0004] The purpose of this invention is to provide a power grid stability control method based on real-time risk assessment. This method can effectively solve the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution:

[0006] A power grid stability control method based on real-time risk assessment includes the following specific steps:

[0007] Step (1) Real-time acquisition of multi-source power grid operation data: Real-time acquisition of voltage amplitude, frequency, frequency change rate, switch status and fault current characteristic data of each key node in the entire network through synchronous phasor measurement unit, smart meter and protection device, and time alignment and data verification to form a multi-dimensional operation status vector under a unified time reference.

[0008] Step (2) Constructing a dynamic risk assessment model: Based on the multidimensional operating state vector, the system's comprehensive risk index is calculated using the following risk assessment function:

[0009]

[0010] in For a moment The comprehensive risk index, This represents the total number of nodes in the power grid participating in the evaluation. For the first The weight coefficients of the nodes, It is a non-linear activation function. For the first Node at time The offset relative to the rated voltage, For the first Node frequency at time The instantaneous rate of change, For the first Node at time Fault current characteristic quantity;

[0011] Step (3) Identify key weak links and risk propagation paths: Based on the comprehensive risk index and its spatial distribution, combined with the power grid topology, calculate the risk transmission sensitivity of each branch, identify nodes and connecting branches whose risk values ​​exceed the preset threshold, and form a dynamic risk hotspot map;

[0012] Step (4) Generate adaptive control instructions: Based on the magnitude and growth trend of the comprehensive risk index, establish a mapping relationship between risk level and control action intensity, and dynamically adjust the execution amount and priority of control measures such as machine tripping, load shedding, reactive power compensation or line switching;

[0013] Step (5) Execute closed-loop feedback control: Send the control command to the execution unit and continuously monitor the operating status of the system after control, update the risk assessment results, and if the risk index does not drop below the safety threshold, iteratively optimize the control strategy until the system returns to a stable state.

[0014] Preferably, in step (1), the sampling frequency of the synchronous phasor measurement unit is not lower than the preset frame rate, the time synchronization accuracy is better than the preset time threshold, and the collected voltage amplitude and phase angle data are used for subsequent calculations after low-pass filtering and anti-aliasing processing to ensure the high fidelity of the data in the transient process.

[0015] Preferably, the nonlinear activation function in step (2) The S-type sigmoid function is used, and its expression is: Where x is the risk contribution value of each node. This is the steepness coefficient, with a value within a predetermined range, to enhance sensitivity to critical risk states; node weight coefficient. The value is determined by weighting the nodes based on their electrical distance in the network, load importance, and historical failure frequency, and the sum is normalized to 1.

[0016] Preferably, the voltage offset in step (2) Defined as ,in Rated voltage; frequency change rate Real-time estimation using the five-point central difference method, with a window length of a predetermined time period; fault current characteristic quantities. This is the effective value of the zero-sequence current or negative-sequence current, and is set to 0 when there is no fault.

[0017] Preferably, in step (3), the risk transmission sensitivity is calculated by the local linearization method of the Jacobian matrix, reflecting the degree of influence of a certain node's risk disturbance on the risk value of adjacent nodes; the risk hotspot map is dynamically displayed on the scheduling and monitoring interface in the form of a heat map, and the update cycle is no greater than the preset time threshold.

[0018] Preferably, in step (4), the risk level is divided into multiple levels: when When the value is below the first preset threshold, it is in a normal state and no control is triggered; when... When the value is between the first preset threshold and the second preset threshold, a warning state is activated, and reactive power compensation adjustment is initiated; when... When the load is between the second and third preset thresholds, it is an emergency state, and secondary loads are cut off according to priority; when... When the value is not less than the third preset threshold, it is a critical state, and the machine is disconnected from the main line and the control is executed.

[0019] Preferably, in step (4), the intensity of the controlled action and the risk index are positively correlated non-linearly, and the specific mapping relationship is as follows: ,in To control the quantity, For the maximum permissible control quantity, As the trigger threshold, To regulate the index, the value is set within a predetermined range to avoid overly aggressive control measures.

[0020] Preferably, in step (5), the iteration period of the closed-loop feedback control does not exceed a preset time threshold. The risk index is recalculated and the control effect is evaluated in each iteration. If multiple iterations are performed consecutively... If the decrease is less than the preset percentage, the current control strategy is deemed to be ineffective, and the system will automatically switch to the backup control plan.

[0021] Preferably, the method is integrated into the power grid energy management system, interacts in real time with the state estimation, power flow calculation and security analysis modules, shares the same data bus, and the communication delay is less than a preset time threshold, ensuring a high degree of synchronization between control decisions and the actual state of the system.

[0022] Preferably, the method supports multi-regional collaborative control, with each regional subnet independently calculating its local risk index and exchanging risk information through boundary nodes. The global coordinator generates cross-regional control commands based on the weighted average risk value to achieve wide-area stable control.

[0023] The beneficial effects of this invention are:

[0024] 1. Achieve precise and stable risk-driven control

[0025] By introducing a dynamic risk assessment function that incorporates voltage offset, frequency change rate, and fault current characteristics, a quantifiable and calculable power grid instability probability index is established for the first time, breaking through the traditional control mode based on static thresholds or offline simulation. This method can continuously and in real-time reflect the evolution trend of the system under disturbances, transforming control decisions from "event-triggered" to "risk-driven," significantly improving the foresight and accuracy of control.

[0026] 2. Improve control response speed and adaptive capability

[0027] Thanks to high-frequency data acquisition, risk calculation, and closed-loop feedback mechanisms, this invention can complete the entire process from risk identification to control execution in a very short time, far faster than the response speed of traditional stable control systems. Simultaneously, the nonlinear mapping relationship between control action intensity and risk index ensures that appropriate control intensity is applied at different risk levels, avoiding over-intervention or under-control and enhancing system robustness.

[0028] 3. Enhance adaptability to complex power grid structures

[0029] By integrating grid topology information with risk transmission sensitivity analysis, this invention can accurately identify key weak links and risk propagation paths, making it particularly suitable for strongly coupled and complex grid scenarios such as high-proportion renewable energy integration and multiple DC feeds. The multi-regional collaborative control architecture further expands the method's applicability, supporting wide-area stability coordination of interconnected regional grids.

[0030] 4. Enhance the safety margin and economy of system operation.

[0031] The combination of dynamic risk hotspot maps and hierarchical control strategies enables dispatchers to intuitively grasp vulnerable areas of the system and take mild adjustment measures in the early stages of risk accumulation, effectively delaying or preventing the onset of emergency control. This not only improves power supply reliability but also reduces unnecessary load shedding losses, balancing safety and economic objectives. Attached Figure Description

[0032] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0034] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0035] In the power grid stability control method based on real-time risk assessment of this invention, step (1) involves real-time acquisition of multi-source operation data of the power grid. This is achieved through synchronous acquisition of core electrical quantities reflecting the dynamic behavior of the system via synchronous phasor measurement units (PMUs), smart meters, and protection devices deployed at key nodes throughout the network. Specifically, the sampling frequency of the synchronous phasor measurement units is no less than 50 frames per second, and the time synchronization accuracy is better than 1 microsecond, ensuring that the acquired voltage amplitude and phase angle data have strict temporal consistency in space, providing a high-fidelity foundation for subsequent transient process analysis. The acquired raw data stream is first processed by low-pass filtering to suppress high-frequency noise interference, and then subjected to anti-aliasing processing to prevent spectral aliasing distortion caused by insufficient sampling rate. Simultaneously, the smart meters provide active power, reactive power, and cumulative power data for each load node, with an update cycle of 100 milliseconds; the protection devices report switch status (open / closed) and fault current characteristic data in real time, including the fault occurrence time, duration, and current waveform segments. All data sources carry precise timestamps and are time-aligned using the IEEE 1588 Precise Time Protocol (PTP) or the Global Positioning System (GPS) to form a multi-dimensional operational state vector under a unified time reference. Each dimension of this vector corresponds to a specific physical quantity, such as the... Voltage amplitude at each node ,frequency Switch status The data structure uses a fixed-length binary format stored in a high-speed circular buffer to support millisecond-level real-time reading. The data verification mechanism includes parity check, cyclic redundancy check (CRC), and reasonable boundary check based on historical data. Any data point judged as abnormal or invalid will be marked and repaired by neighboring valid data points through linear interpolation, thereby ensuring the integrity and reliability of the data input to the risk assessment model.

[0036] In the above-mentioned power grid stability control method based on real-time risk assessment, step (2), constructing a dynamic risk assessment model, is the core technical aspect of this invention. This model, based on the multi-dimensional operating state vector formed in step (1), uses a comprehensive risk assessment function to quantify the overall instability risk level of the system at any time t. The mathematical expression of this function is as follows:

[0037]

[0038] in For a moment The comprehensive risk index, This represents the total number of nodes in the power grid participating in the evaluation. For the first The weight coefficients of the nodes, It is a non-linear activation function. For the first Node at time The offset relative to the rated voltage, For the first Node frequency at time The instantaneous rate of change, For the first Node at time The fault current characteristic quantity. Specifically, voltage offset. Defined as ,in For the system's rated voltage (e.g., 220kV or 500kV), this normalization process eliminates dimensional differences between nodes at different voltage levels. Frequency change rate It is not obtained through direct measurement, but rather estimated in real time by applying the five-point central difference method to five consecutive frequency sampling points. The calculation window length is fixed at 20 milliseconds to achieve a balance between response speed and numerical stability. Fault current characteristic quantity This value is set to 0 during normal system operation; once the protection device detects a fault, it is assigned the effective value (RMS) of either the zero-sequence current or the negative-sequence current during the fault period, selected based on the fault type—zero-sequence current is preferred for ground faults, while negative-sequence current is used for phase-to-phase faults. (Nonlinear activation function) The sigmoid function is used, and its expression is: ,in, The steepness coefficient k ranges from 2 to 5, representing the risk contribution value for each node. The choice of this parameter is crucial; a larger k value results in a steeper slope of the function near x=0, significantly enhancing the model's sensitivity to critical risk states (i.e., x approaching 0) and making the risk index... It can rapidly improve when the system state experiences a small but dangerous shift. Node weight coefficients. Instead of uniform distribution, the distribution is weighted based on three dimensions: the electrical distance of the node in the network, the importance level of the load it supplies (e.g., primary, secondary, and tertiary loads), and the node's historical failure frequency. Ultimately, all... The sum is normalized to 1 to ensure that the range of values ​​for R(t) has a clear physical meaning. Coefficients , , The initial values ​​of the sensitivity gains for each risk factor are calibrated through offline simulation and can be fine-tuned online during system operation based on the actual control effect.

[0039] In the aforementioned power grid stability control method based on real-time risk assessment, step (3) identifies key weak links and risk propagation paths, aiming to transform the abstract comprehensive risk index R(t) into an operational, spatialized risk insight. This step first transforms the risk contribution value of each node calculated in step (2) (i.e., The risk assessment function R(t) is mapped to its physical location in the power grid topology diagram. The power grid topology is stored in memory as an adjacency matrix, where the matrix elements represent the electrical connections between nodes and branch impedance parameters. Based on this, the risk transmission sensitivity of each branch is calculated, which is obtained through the local linearization method of the Jacobian matrix. Specifically, the partial derivatives of the risk assessment function R(t) with respect to the state variables of each node are taken to construct a local Jacobian matrix J, whose elements are... Reflects the first The state disturbance of the nth node affects the nth node. The impact of risk values ​​at each node is assessed. By analyzing the eigenvalues ​​and eigenvectors of this matrix, the dominant patterns and key channels of risk propagation in the system can be identified. Nodes with risk values ​​exceeding a preset threshold (e.g., 0.6) are marked as "risk hotspots," and the branches connecting these hotspots are identified as "risk propagation paths." All this information is integrated to generate a dynamic risk hotspot map, which is rendered in real-time on the dispatcher's monitoring interface in the form of a heatmap. The colors transition from green (low risk) through yellow and orange to red (high risk), visually displaying the spatial distribution of vulnerable areas in the system. This heatmap is updated no more than 100 milliseconds, ensuring that the dispatcher can grasp the evolution of the system's risk situation in near real-time.

[0040] In the aforementioned power grid stability control method based on real-time risk assessment, step (4) generates adaptive control commands, establishing a direct mapping from risk perception to control execution. This step first considers the comprehensive risk index... The magnitude of the value and its first derivative (i.e., the growth trend) are used to classify the system state. The risk level is divided into four levels: when... When 0.3 < 0.3, the system is in normal condition and does not trigger any control actions; when 0.3 ≤ When the voltage is less than 0.6, the system enters an early warning state. At this time, reactive power compensation regulation is activated, which smooths voltage fluctuations by switching capacitor banks or adjusting the output of the static var compensator (SVC). When 0.6 ≤ When the value is less than 0.85, the system is in an emergency state, and the control strategy is upgraded to disconnect secondary loads according to preset priorities. The priority list is pre-configured based on the importance, interruptibility, and geographical location of the loads. When the value is ≥ 0.85, the system enters a critical state, requiring generator disconnection and main line disconnection control to prevent the instability range from expanding. The intensity of the control action is not a simple step-like change, but rather exhibits a non-linear positive correlation with the risk index, with the specific mapping relationship as follows:

[0041]

[0042] in, For control quantities to be issued (such as load power to be cut off or generator output). This is the maximum permissible control quantity corresponding to this control measure. The trigger threshold for this control measure (e.g., for load shedding, =0.6), where η is the control index, ranging from 1.2 to 1.8. This nonlinear relationship is designed to avoid overly aggressive control actions. Just over When the increment of u(t) is small, a mild intervention is applied; as The sharp rise, The growth rate is accelerating to match the rapid deterioration of risks. Furthermore, the prioritization of control measures is not only based on risk level but also dynamically considers currently available control resources, the health status of execution units, and geographical distribution to ensure the feasibility and effectiveness of control instructions.

[0043] In the above-mentioned power grid stability control method based on real-time risk assessment, step (5) executes closed-loop feedback control, forming a complete and adaptive control loop. This step first sends the adaptive control command generated in step (4) to the corresponding execution unit, including circuit breakers, load switches, and reactive power compensation device controllers, through a high-speed communication network (such as IEC 61850 GOOSE messages). After the command is sent, the system immediately enters the monitoring mode, continuously collects the operating status data of the system after control, and re-executes steps (1) to (4) in the next iteration cycle (not exceeding 200 milliseconds) to calculate a new comprehensive risk index. If the newly calculated If the value has dropped below the safety threshold (e.g., 0.3), the control is considered successful, and the system exits the closed-loop control mode. If... If the temperature does not drop below the safety threshold, the system will iteratively optimize the control strategy, such as adjusting the control input. The size of the control, switching to a higher priority control measure, or activating a backup control plan. Specifically, if after three consecutive iterations... If the decrease in all parameters is less than 5%, the system determines that the current control strategy has failed. Possible causes include execution unit failure, model parameter mismatch, or a continuous increase in the disturbance source. In this case, the system will automatically switch to a preset, more conservative backup control plan to ensure the system's safety baseline. The entire closed-loop feedback control process is deeply integrated with the power grid energy management system (EMS), sharing the same high-speed data bus. It interacts with the state estimation, power flow calculation, and security analysis modules in real time, with communication latency strictly controlled within 10 milliseconds, thereby ensuring that control decisions are always highly synchronized with the actual system state.

[0044] Based on the description of the first stage method in Embodiment 1 above, a specific application example is now constructed to further illustrate the implementation details and effects of the present invention. Assume that a regional power grid includes a 500kV main grid and several 220kV subgrids, connected to a large number of wind power and photovoltaic power stations, with low system inertia and susceptible to disturbances. At a certain moment, a 500kV transmission line experiences a momentary fault due to a lightning strike, the protection device operates correctly, but after the fault is cleared, the system experiences low-frequency oscillations. The control system of the present invention is immediately activated: in step (1), the PMU captures the voltage drops of multiple 220kV nodes at a frequency of 50 frames / second ( (reaching 0.15 pu) and frequency dropping rapidly ( (Up to -1.2 Hz / s); in step (2), the risk assessment model calculates Within 100 milliseconds, it spiked from 0.25 to 0.72; in step (3), the risk hotspot map clearly showed that the oscillation center was located on the connecting line between two large wind farms; in step (4), the system was determined to be in an emergency state, based on the nonlinear mapping relationship. The system calculates the 50MW of secondary industrial load that needs to be cut off and generates control commands specific to the substation and feeder. In step (5), the commands are issued and executed within 150 milliseconds, and the system re-evaluates the situation within the next 200 milliseconds. The reactive power dropped to 0.45, and the system continued to perform a round of mild reactive power compensation, eventually reducing it within 500 milliseconds. The value stabilized at 0.28, successfully preventing system crashes. This example fully demonstrates the significant advantages of this invention in terms of response speed, control precision, and adaptive capability.

[0045] Furthermore, the method of this invention supports multi-regional collaborative control. In the scenario of interconnected power grids across large regions, each regional subgrid can independently run all five steps of this invention to calculate the local comprehensive risk index. Each region exchanges local risk information, including the risk value and gradient of the boundary nodes, through its boundary nodes (i.e., nodes connected to other regions). A global coordinator receives boundary risk information from all regions and, based on factors such as the size and importance of each region, adjusts the risk information accordingly. The weighted average is used to obtain the global risk index. .when When an indication is given that there is a risk of cross-regional risk propagation, the global coordinator will generate cross-regional control instructions, such as coordinating two regions to simultaneously provide reactive power support or orderly load shedding, thereby achieving wide-area stability control and effectively addressing complex stability issues such as weakly damped oscillations between regions.

[0046] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A power grid stability control method based on real-time risk assessment, characterized in that, Includes the following steps: S1: Real-time acquisition of multi-source operation data of the power grid, obtaining voltage amplitude, frequency, frequency change rate, switch status and fault current characteristic data of each key node in the entire network through synchronous phasor measurement unit, smart meter and protection device, and performing time alignment and data verification to form a multi-dimensional operation status vector under a unified time reference; S2: Construct a dynamic risk assessment model. Based on the multi-dimensional operating state vector, use a risk assessment function to calculate the system's comprehensive risk index. The risk assessment function is a weighted sum of the risk contribution values ​​of each node. The risk contribution value of each node is obtained by a linear combination of the voltage offset, frequency change rate, and fault current characteristic quantity of the node applied by the nonlinear activation function. S3: Identify key weak links and risk propagation paths. Based on the spatial distribution of the comprehensive risk index and the power grid topology, calculate the risk transmission sensitivity of each branch, identify nodes and connecting branches whose risk values ​​exceed the preset threshold, and generate a dynamic risk hotspot map. S4: Generate adaptive control instructions, establish a mapping relationship between risk level and control action intensity based on the magnitude and growth trend of the comprehensive risk index, and dynamically adjust the execution amount and priority of machine tripping, load shedding, reactive power compensation or line switching. S5: Execute closed-loop feedback control, send the control command to the execution unit, continuously monitor the operating status of the system after control, update the risk assessment results, and if the risk index does not drop below the safety threshold, iteratively optimize the control strategy until the system returns to a stable state.

2. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The sampling frequency of the synchronous phasor measurement unit is no less than 50 frames per second, and the time synchronization accuracy is better than 1 microsecond. The collected voltage amplitude and phase angle data are used for subsequent calculations after low-pass filtering and anti-aliasing processing.

3. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The nonlinear activation function is the sigmoid function, whose expression is: ,in, The risk contribution value for each node, the steepness coefficient k ranges from 2 to 5; the weight coefficient of the node is determined by weighting the node's electrical distance in the network, load importance and historical failure frequency, and the sum of the weight coefficients of all nodes is normalized to 1.

4. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The voltage offset is defined as the absolute value of the difference between the node voltage amplitude and the rated voltage divided by the rated voltage; the frequency change rate is estimated by the five-point center difference method for five consecutive frequency sampling points, with a calculation window length of 20 milliseconds. The fault current characteristic is set to 0 when there is no fault, takes the effective value of zero-sequence current when a ground fault occurs, and takes the effective value of negative-sequence current when a phase-to-phase fault occurs.

5. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The risk transmission sensitivity is constructed by taking the partial derivative of the risk assessment function with respect to the state variables of each node to construct a local Jacobian matrix, and the degree of risk disturbance impact between nodes reflected by the elements of the matrix is ​​analyzed; the dynamic risk hotspot map is displayed in the form of a heat map on the scheduling and monitoring interface, and the color changes from green, yellow, orange to red from low to high according to the risk value, with an update cycle of no more than 100 milliseconds.

6. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The risk levels are divided into four levels: a comprehensive risk index below 0.3 is a normal state and no control is triggered; a level between 0.3 and 0.6 is a warning state and reactive power compensation adjustment is initiated; a level between 0.6 and 0.85 is an emergency state and secondary loads are cut off according to priority; a level not less than 0.85 is a critical state and the generator tripping and main line disconnection control are executed.

7. The power grid stability control method based on real-time risk assessment according to claim 6, characterized in that, The intensity of the control action is non-linearly positively correlated with the comprehensive risk index, and the control quantity... From the formula It is confirmed that, among them, As a comprehensive risk index, For the maximum permissible control quantity, The control index η ranges from 1.2 to 1.8 to correspond to the trigger threshold of the control measures.

8. The power grid stability control method based on real-time risk assessment according to claim 1, characterized in that, The iteration cycle of the closed-loop feedback control does not exceed 200 milliseconds. In each iteration, the comprehensive risk index is recalculated and the control effect is evaluated. If the decrease in the comprehensive risk index is less than 5% after three consecutive iterations, the current control strategy is determined to be ineffective and the system is automatically switched to the backup control plan. The method is integrated into the power grid energy management system and shares the same data bus with the state estimation, power flow calculation and security analysis modules, with a communication delay of less than 10 milliseconds.

Images