Islanding event stability and coordination control method based on off-site load transfer

Abstract

The application discloses a power grid isolated network event stability coordination control method based on load remote power transfer, comprising the following steps: collecting power grid operation states to construct isolated network state feature set, identifying risk period and weak section set according to isolated network power instability critical entropy criterion algorithm, starting cross-section power transfer coordination balancing algorithm to generate power transfer operation sequence, calling load remote allocation supply and demand dynamic balancing model to solve the coordinated allocation scheme of load reduction at the receiving end and unit output increase at the sending end, embedding the allocation scheme into the isolated network transient stability double-layer constraint accounting platform to obtain the feasible regulation and control strategy through double-layer iteration of frequency voltage dynamic constraint verification and power angle stability constraint verification, and issuing the pre-control instruction and post-stability control instruction to the automatic device according to the feasible regulation and control strategy. The method realizes the whole process regulation and control of the isolated network operation state through the cooperation of pre-control and post-stability.

Description

Technical Field

[0001] This invention relates to the field of collaborative control technology for isolated grid events, and in particular to a method for coordinated control of isolated grid events based on load transfer to other locations. Background Technology

[0002] Isolated grid operation refers to the operating state where a local power system supplies power independently after being disconnected from the main grid. In this state, the system inertia is significantly reduced and the frequency and voltage regulation capabilities are weak. Once a power disturbance occurs, it is very easy to cause frequency collapse or voltage instability. With the large-scale integration of new energy sources and the increasingly complex AC / DC hybrid grid structure, the receiving-end grid is facing more and more fault modes such as simultaneous blocking of multiple DC lines and unplanned disconnection of inter-regional tie lines. The frequency and impact range of isolated grid operation scenarios are both on the rise, and there is an urgent need to build an isolated grid stability control technology system that can adapt to complex operating conditions.

[0003] Existing technologies mainly employ two implementation paths. The first is pre-planned control based on offline simulation and pre-calculation. This involves using time-domain simulation software to perform batch calculations on typical fault sets, generating a mapping table between faults and control measures. When an actual fault occurs, the table is consulted based on the fault type to invoke the preset strategy. The second is local control based on local measurement response. This utilizes the frequency change rate or voltage drop depth to trigger automatic devices such as low-frequency load shedding and low-voltage load shedding. Load shedding is implemented according to pre-set cycles and load shedding amounts, while simultaneously combining the unit's primary frequency regulation capability to suppress frequency deviation.

[0004] The aforementioned existing technologies have two main drawbacks. First, offline contingency plans cannot cover the continuous variation of fault modes. When the actual fault characteristics deviate from the preset scenario, the adaptability of the preset strategies drops sharply. Furthermore, the control measures lack a coordinated consideration of the cross-sectional transmission margin and the load transfer capability, which can easily lead to local overload or waste of control resources. Second, local control relies on triggering by a single electrical quantity threshold. There is a lack of global information interaction and action timing coordination between devices. In isolated network scenarios with multi-section power coupling, this can easily lead to over-adjustment or under-adjustment of control measures. After multiple actions, the system state deviates from the expected trajectory, making it difficult to achieve an organic connection between pre-control and post-operational stability. Summary of the Invention

[0005] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides a method for coordinated control of power grid isolated events based on load transfer to different locations.

[0006] The technical solution adopted in this invention is a method for coordinated control of isolated grid events based on load transfer between different locations, comprising the following steps: S1, collecting topology connection status, generator active power output range, bus voltage amplitude and phase angle, and cross-sectional transmission power limit from the grid operation and dispatch layer to construct an isolated grid operation status feature set; S2, performing transient stability margin quantitative assessment on the status feature set according to the isolated grid power instability critical entropy criterion algorithm, and identifying the risk period and weak cross-section set that power fluctuations may induce isolated grid collapse; S3, initiating a cross-section power transfer coordinated balancing algorithm for the weak cross-section set, calculating the transferable power of each tie line. S4. Based on the power amplitude and priority of the transfer path, a transfer operation sequence is generated; S5. The load off-site allocation supply and demand dynamic balance model is invoked to solve the coordinated allocation scheme of the load reduction at the receiving end and the output increase of the sending end unit with the transfer power as input; S6. The allocation scheme is embedded into the islanded grid transient stability dual-layer constraint calculation platform. The upper layer performs frequency and voltage dynamic constraint verification, and the lower layer performs power angle stability constraint verification. The feasible control strategy that satisfies all constraints is obtained through dual-layer iteration; S7. Based on the feasible control strategy, pre-control instructions and post-stability control instructions are issued to the automatic device to carry out the whole process control of the islanded grid operation status.

[0007] Furthermore, the expression for the islanded network power instability critical entropy criterion algorithm is as follows:

[0008] ,

[0009] in, The cumulative entropy value for power imbalance. This represents the start time of the disturbance. This is the critical resection time. Let be the instantaneous deviation between the total generating power and the total load power within the isolated grid at time t. This is the reference value for the rated power of an isolated grid;

[0010] ,

[0011] in, As an instability risk factor, To preset the entropy threshold, The attenuation coefficient is... The time rate of change of power deviation at the critical cutoff moment; when If the value is greater than 1, the current operating state is considered to have a risk of instability.

[0012] Furthermore, the cross-section power transfer collaborative equalization algorithm is expressed as follows:

[0013] ,

[0014] in, This represents the total power that can be transferred from the i-th weak section. Let i be the set of connecting lines adjacent to the i-th cross section. These are the voltage amplitudes of the busbars on both sides of the cross-section. For the line The reactance value, These are the phase angles of the bus voltages on both sides of the cross-section;

[0015] ,

[0016] in, This is the cumulative amount of supply imbalance during all time periods. The duration of the scheduling cycle. This represents the total number of weak cross sections. The transmission power limit for the i-th section, For the integration time variable; by minimizing Determine the optimal power transfer ratio for each section.

[0017] Furthermore, the dynamic balance model for load redistribution and supply-demand equilibrium is expressed as follows:

[0018] ,

[0019] in, To account for the net power difference of the isolated grid after the adjustment, As the reference voltage, Operating voltage This is the partial derivative of the load power with respect to the voltage. This represents the total number of adjustable generating units. This is the output adjustment amount for the k-th unit. To reduce the total number of load nodes, Let m be the original load reduction amount for the m-th load node. This is the load frequency regulation coefficient. This is the frequency deviation. For the rated frequency; make The allocation scheme that approaches zero is the solution for dynamic supply and demand balance.

[0020] Furthermore, the isolated network transient stability dual-layer constraint calculation platform includes a hybrid architecture computing unit composed of a digital signal processor and a field-programmable gate array. In the upper-layer frequency and voltage dynamic constraint verification, the frequency deviation iteration convergence threshold is set to 0.02, and the voltage deviation iteration convergence threshold is set to 0.01. In the lower-layer power angle stability constraint verification, the maximum allowable deviation of the relative power angle swing curve is set to 0.15. The global convergence judgment condition of the dual-layer iteration is that the weighted sum of the squares of the upper-layer verification residual and the lower-layer verification residual is less than 0.005. The isolated network transient stability dual-layer constraint calculation platform reads the allocation scheme parameters cyclically at a 1-millisecond cycle and outputs the feasible control strategy identification code marked by the dual-layer verification.

[0021] Further, S2 includes the following sub-steps: S21, extracting the active power output value of each generator and the load power value of each node from the state feature set, and calculating the instantaneous power deviation sequence of the isolated network; S22, using the power deviation sequence as input, approximating the cumulative entropy value of power imbalance using the trapezoidal integral method; S23, generating an instability risk factor based on the ratio of the cumulative entropy value to a preset entropy threshold value, combined with the power deviation time change rate; S24, marking the time when the instability risk factor is greater than 1 as the risk period, and simultaneously locating the cross-sections where the transmission power exceeds the limit during the risk period as the set of weak cross-sections.

[0022] Further, S3 includes the following sub-steps: S31, for each section in the set of weak sections, read the voltage amplitude and phase angle data of the busbars at both ends of its tie line, and calculate the single-line transferable power value of each tie line; S32, collect the transferable power values ​​of the tie lines belonging to the same section, and generate the total transferable power of the section; S33, calculate the cumulative amount of power transfer imbalance throughout the time period based on the degree of deviation between the total transferable power of each section and the transmission power limit; S34, with minimizing the cumulative amount of imbalance as the optimization objective, solve the optimal allocation coefficient of the power transfer of each section by gradient descent method, and generate the power transfer operation sequence by arranging the coefficients in descending order.

[0023] Further, S4 includes the following sub-steps: S41, read the voltage operating value and reference value of the receiving-end node involved in the power transfer operation sequence, and perform voltage-sensitive reduction pre-calculation on the receiving-end load power; S42, read the current upper and lower limits of the output of the adjustable units at the sending end, and determine the feasible range of the output increase of each unit; S43, substitute the load reduction and unit increase into the supply and demand dynamic balance equation, and calculate the net difference of the isolated grid power after allocation; S44, use the Newton-Raphson method to iteratively adjust the numerical ratio of the reduction and increase until the absolute value of the net difference is lower than the convergence threshold, and output the current ratio as the coordinated allocation scheme.

[0024] Further, S5 includes the following sub-steps: S51, mapping the coordinated allocation scheme to frequency response curves and voltage response curves, and the upper-level calculation platform performs dynamic constraint verification of frequency deviation peak value and voltage drop depth; S52, if the upper-level verification passes, the scheme is transferred to the lower-level calculation platform, and the lower-level platform performs stability constraint verification of the generator power angle relative swing curve; S53, if the lower-level verification fails, the power angle instability indicator is fed back to the upper-level platform, and the upper-level platform corrects the power allocation weight in the allocation scheme accordingly; S54, repeating the iterative correction process between the upper and lower levels until the allocation scheme simultaneously meets the frequency and voltage dynamic constraints and the power angle stability constraints, and marking the scheme as a feasible control strategy.

[0025] On the other hand, this invention also discloses a grid islanding event stability and coordinated control system based on load transfer between different locations, comprising: an islanding operation status feature acquisition unit, whose input is connected to the grid dispatch data network to obtain topology connection status, generator output range, bus voltage phasor and cross-sectional power limit, and whose output is to output a set of status features; an islanding instability risk identification unit, whose input is connected to the output of the feature acquisition unit, and which internally embeds an islanding power instability critical entropy criterion algorithm to output risk period identifiers and a set of weak cross-sections; and a cross-section transfer strategy generation unit, whose input is connected to the output of the risk identification unit, and which internally embeds a cross-section transfer power coordinated balancing algorithm to generate a transfer operation sequence and output it to the load dispatching unit. The load off-site allocation supply and demand balance solution unit has its input terminals connected to the output terminal of the transfer strategy generation unit and the load parameter port of the dispatch data network, respectively. It internally stores the load off-site allocation supply and demand dynamic balance model to output a coordinated allocation scheme. The transient stability dual-layer constraint verification unit has its input terminal connected to the output terminal of the supply and demand balance solution unit. It internally integrates the frequency and voltage dynamic constraint verification module and the power angle stability constraint verification module. After dual-layer iterative calculation, it outputs a feasible control strategy. The pre-control and stability command output unit has its input terminal connected to the output terminal of the verification unit and its output terminal connected to the automatic generation control device and the stability control device. It breaks down the feasible control strategy into pre-control parameters and post-stability action parameters and issues them separately.

[0026] The present invention has the following beneficial effects:

[0027] By constructing a feature set of isolated grid operation status and invoking the critical entropy criterion algorithm for isolated grid power instability, the risk periods and weak sections induced by power fluctuations can be continuously identified. This solves the defect that offline contingency plans cannot cover the continuously changing space of faults, making risk identification no longer dependent on preset fault templates. A cross-section power transfer collaborative balancing algorithm is used to globally optimize the allocation of transferable power across multiple sections. Combined with a dynamic balance model of load dispatching and supply and demand, the collaborative ratio of receiving-end reduction and sending-end increase is solved, achieving balanced utilization of control resources across multiple sections and effectively avoiding local overload and waste of control resources. Furthermore, a dual-layer constraint accounting platform for isolated grid transient stability is used to dynamically apply frequency and voltage constraints to the dispatching scheme. The dual-layer iterative verification of constraints and power angle stability constraints ensures that the pre-control strategy itself satisfies multiple stability boundaries. At the same time, the feasible control strategy is broken down into pre-control commands and post-stability control commands, which are issued to the automatic generation control device and the stability control device respectively. This allows the pre-control to actively adjust the isolated grid operation base point to reduce the risk of instability, while the post-stability control only intervenes as a supplementary means when the pre-control is insufficient to completely suppress disturbances. The two layers of control form a clear timing coordination and functional complementarity relationship, thereby solving the over-adjustment or under-adjustment problem caused by local control relying on a single electrical quantity threshold and lacking global coordination. This achieves adaptive matching and effective expansion of stability boundaries in the whole process of grid isolated operation control. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention.

[0029] Figure 2 This is a flowchart of method step S2 of the present invention;

[0030] Figure 3 This is a flowchart of method step S3 of the present invention;

[0031] Figure 4 This is a flowchart of method step S4 of the present invention;

[0032] Figure 5 This is a flowchart of step S5 of the method of the present invention. Detailed Implementation

[0033] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] like Figure 1 As shown, the method for coordinated control of isolated grid events based on load transfer includes the following steps: S1, collecting topology connection status, generator active power output range, bus voltage amplitude and phase angle, and cross-sectional transmission power limit from the grid operation and dispatch layer to construct an isolated grid operation status feature set; S2, performing transient stability margin quantitative assessment on the status feature set according to the isolated grid power instability critical entropy criterion algorithm, identifying the risk period and weak cross-section set that power fluctuations may induce isolated grid collapse; S3, initiating a cross-section power transfer coordinated balancing algorithm for the weak cross-section set, calculating the transferable power amplitude of each tie line and S4. Prioritize the transfer path and generate a transfer operation sequence; S5. Call the load off-site allocation supply and demand dynamic balance model, and use the transfer power as input to solve the coordinated allocation scheme of the load reduction at the receiving end and the output increase of the sending end unit; S6. Embed the allocation scheme into the islanded grid transient stability dual-layer constraint calculation platform. The upper layer performs frequency and voltage dynamic constraint verification, and the lower layer performs power angle stability constraint verification. After dual-layer iteration, obtain the feasible control strategy that satisfies all constraints; S7. Based on the feasible control strategy, issue pre-control instructions and post-stability control instructions to the automatic device to carry out the whole process control of the islanded grid operation status.

[0035] Step S1 involves collecting topology connection status, generator active power output range, bus voltage amplitude and phase angle, and cross-sectional transmission power limit from the power grid operation and dispatch layer to construct an isolated grid operation status feature set. This step is executed in the data acquisition and monitoring control system of the power grid dispatch center, reading electrical measurement data of the entire power grid from the real-time database of the energy management system at a period of 100 milliseconds. Topology connection status includes the open / closed position identifiers of each circuit breaker and disconnector; generator active power output range records the current actual output and adjustable upper and lower limits of each online unit; bus voltage amplitude is recorded in per-unit form with 4 decimal places of precision; voltage phase angle is recorded in degrees; and cross-sectional transmission power limit refers to the maximum active power value allowed to pass through tie lines or transformer branches under thermal stability constraints. All collected data are uniformly encapsulated into structured state feature vectors according to time-series labels and stored in a high-speed cache for subsequent steps. During the data acquisition process, each measurement point is marked with an over-limit flag. Operating states where the voltage amplitude deviates from the range of 0.95 to 1.05 per unit and the cross-sectional power exceeds the limit by 90% are marked as warning states. These states, along with the corresponding cross-sectional flag and bus node number, are entered into the status feature set as the initial input base for subsequent risk assessment.

[0036] Step S2 involves quantitatively assessing the transient stability margin of the state feature set based on the islanded grid power instability critical entropy criterion algorithm, identifying risk periods and vulnerable sections where power fluctuations could induce islanded grid collapse. This step extracts the generator active power output sequence and the load power sequence of each node from the state feature set in the high-speed buffer, calculating the instantaneous deviation between the total generator power and the total load power within the islanded grid point by point, using the actual sampling point interval as the time step. The instantaneous deviation sequence is input into the power imbalance cumulative entropy calculation module, which uses a numerical integration method to progressively accumulate the entropy contribution of the deviation value along the time axis. After the cumulative entropy value is calculated, it is compared with the entropy threshold pre-stored in the system, which is calibrated based on the critical stability boundary in the historical power grid operating data. Simultaneously, the time rate of change of the power deviation at the current moment is calculated, and the ratio of the cumulative entropy value to the entropy threshold is multiplied by the exponential decay term of this rate of change to obtain the instability risk factor. When the instability risk factor is greater than 1, it is determined that the moment is in a risk state. All time intervals of continuous risk states are merged into a risk period. During the risk period, the deviation of the transmission power value from the preset limit is scanned section by section. Sections with a deviation of more than 15% are classified into a weak section set. Each section in the set records the bus number on both sides and the current transmission power value.

[0037] Step S3: For the set of weak cross sections, initiate the cross-section power transfer collaborative balancing algorithm, calculate the transferable power amplitude and transfer path priority of each tie line, and generate a transfer operation sequence. This step first reads the bus voltage amplitude and phase angle data on both sides of each cross section in the weak cross section set, extracts the line reactance parameters, and calculates the additional active power transmission increment that each tie line under that cross section can carry at the current operating point according to the AC line active power transmission equation. This increment uses the difference between the upper limit and the current value as the transferable power amplitude. The transferable power amplitudes of all tie lines in the same cross section are summed to obtain the total transferable power of that cross section. When multiple weak cross sections exist, the square deviation of the total transferable power of each cross section from the transmission power limit is calculated sequentially. The deviations of all cross sections at each time section are accumulated over the scheduling cycle to obtain the cumulative transfer imbalance over the entire time period. The power transfer ratio of each cross-section is used as the variable to be optimized. The objective function is to minimize the accumulated imbalance. A gradient descent optimization algorithm is used to iteratively solve for the optimal allocation coefficient, with an iteration step size of 0.01 and a maximum number of iterations of 500. After convergence, the cross-sections are arranged in descending order of allocation coefficient. For each cross-section, the tie line with the highest power transfer capability is selected as the preferred power transfer channel. The cross-section number and the preferred channel number are then combined sequentially to form the power transfer operation sequence.

[0038] Step S4 involves calling the load relocation and supply-demand dynamic balance model, using the transferred power as input to solve for the coordinated allocation scheme of load reduction at the receiving end and output increase at the sending end. This step extracts the bus number and corresponding voltage operating value of the receiving-end node from the transfer operation sequence, reads the currently connected load power value of that bus, and pre-calculates the load reduction potential value of that node under different voltage levels based on the sensitivity relationship between load power and voltage. Simultaneously, it reads the upper limit and current output value of the adjustable units in the sending-end region, using the difference between the upper limit and the current value as the feasible upper limit for the output increase of each unit. The load reduction of each node at the receiving end and the output increase of each unit at the sending end are used as variables, and these variables are uniformly substituted into the supply-demand dynamic balance equation. The equation calculates the net power difference value within the isolated grid after allocation. The Newton-Raphson iterative method is used to adjust the variable values ​​iteratively. The convergence criterion for the iteration process is that the absolute value of the net power difference is less than 0.001. In each iteration, the ratio of load reduction to unit output adjustment is guided by the positive or negative direction of the net difference. The iteration terminates when the convergence criterion is met, and the current load node reduction allocation table and unit output adjustment allocation table are combined into a coordinated dispatch scheme output.

[0039] Step S5 involves embedding the dispatching scheme into the islanded grid transient stability dual-layer constraint calculation platform. The upper layer performs frequency and voltage dynamic constraint verification, while the lower layer performs power angle stability constraint verification. Through dual-layer iteration, a feasible control strategy satisfying all constraints is obtained. This step maps each control variable in the coordinated dispatching scheme to the frequency response curve and voltage response curve of the islanded grid after dispatching. The frequency response curve is jointly determined by the generator governor model and load frequency characteristics, while the voltage response curve is jointly determined by the excitation regulator model and load voltage characteristics. The upper-layer calculation platform reads the peak frequency deviation from the frequency response curve and compares it with a threshold value of 0.5 Hz. Simultaneously, it reads the voltage sag depth from the voltage response curve and compares it with a threshold value of 0.1 per unit. If both the peak frequency deviation and the voltage sag depth are within the threshold values, the dynamic constraint verification is considered successful, and the scheme is passed to the lower-layer calculation platform. The lower-layer calculation platform uses a time-domain simulation method to calculate the relative power angle swing curve between each generator rotor, extracts the maximum deviation value of the swing curve over the entire simulation time, and compares it with a 15-degree threshold value. If the maximum deviation is within the threshold value, the power angle stability constraint verification is deemed successful, and a feasible control strategy is directly output. If the lower-level verification fails, a power angle instability flag is fed back to the upper level. Based on this flag, the upper level shifts the power allocation weight in the allocation scheme by 0.05 towards the section farther from the unstable unit. After correction, the dynamic constraint verification is re-executed and transmitted to the lower level again. The two levels iterate repeatedly until all constraints are satisfied simultaneously.

[0040] Step S6 involves issuing pre-control commands and post-stabilization control commands to the automatic control devices based on feasible control strategies, completing the full-process control of the isolated grid operation. This step breaks down the control parameters in the feasible control strategies into two command sets. The pre-control command set includes the upward adjustment of the active power output setpoint of the sending-end units and the opening and closing sequence of the load transfer switches at the receiving end. This set is issued to the unit controllers of the automatic generation control device and the automatic standby power switching devices of the receiving-end substation with a 5-second lead. The post-stabilization control command set includes the operating frequency setting and load reduction of the low-frequency load shedding device for each cycle, and the operating voltage setting and load reduction of the low-voltage load shedding device for each cycle. This set is issued to the frequency and voltage emergency control devices installed at each load node. The execution priority of the pre-control commands is higher than that of the post-stabilization control commands. After receiving the commands, the automatic generation control device gradually increases the unit output at a rate of 10% per second, adjusting the deviation between the current output and the target value. When the islanded network frequency drops to 49.5 Hz, the low-frequency load shedding device performs the first round of load shedding according to the frequency setting value of the first round of operation. If the frequency continues to drop, subsequent rounds are executed in turn. All issued commands are accompanied by timestamps and check codes. After parsing the commands, the receiving device sends back an acknowledgment message to the dispatch center.

[0041] Preferably, the algorithm for determining the critical entropy criterion for power instability in isolated networks is expressed as follows:

[0042] ,

[0043] in, The cumulative entropy value for power imbalance. This represents the start time of the disturbance. This is the critical resection time. Let be the instantaneous deviation between the total generating power and the total load power within the isolated grid at time t. This is the reference value for the rated power of an isolated grid;

[0044] ,

[0045] in, As an instability risk factor, To preset the entropy threshold, The attenuation coefficient is... The time rate of change of power deviation at the critical cutoff moment; when If the value is greater than 1, the current operating state is considered to have a risk of instability.

[0046] Specifically, the construction of the cumulative entropy value for power imbalance in the islanded grid power instability critical entropy criterion algorithm is based on the idea of ​​uncertainty quantification in information entropy theory. The instantaneous power deviation between generators and loads within the islanded grid is considered a characterizing quantity of the system's orderliness. As the absolute value of the power deviation increases, the uncertainty of the system's operating state increases, and the entropy value increases accordingly. This formula uses the ratio of the instantaneous power deviation to the rated power reference value as the core variable for entropy calculation. An integral accumulation operation is performed on the time axis from the start of the disturbance to the critical cutoff time. During the integration process, the ratio is multiplied by the natural logarithm of the ratio and the result is negative for each time micro-element. The contributions of all micro-element elements are accumulated along the time axis to obtain the cumulative entropy value. The rated power reference value is taken as the per-unit base value of the sum of the rated capacities of all online generators within the islanded grid. The second formula uses the ratio of the cumulative entropy value to a preset entropy threshold as the base term. The threshold is tuned based on the statistical average of the cumulative entropy values ​​under historical critical stability conditions of the islanded grid, typically ranging from 0.5 to a fixed value between 2.0. Simultaneously, an exponential decay term for the rate of change of power deviation at the critical cutoff moment is introduced as a correction coefficient. The decay coefficient ranges from 0.1 to 0.5, and its physical meaning is that the more drastic the change in power deviation, the more urgent the system instability trend. The exponential decay term amplifies the risk factor sharply under high rate of change conditions. The two formulas are logically progressive. The first formula accumulates the degree of power imbalance from the time dimension to obtain the entropy value. The second formula, based on the entropy value, introduces a time-domain dynamic characteristic quantity for correction and outputs the final criterion. In implementation, the algorithm is integrated into the risk assessment server of the power grid dispatch center as a function library. The algorithm module reads the latest power deviation sampling value from the real-time database every 20 milliseconds of the power frequency cycle and performs calculations. When the risk factor is greater than 1 for more than 100 milliseconds, the subsequent control process is triggered.

[0047] Preferably, the cross-section power transfer collaborative equalization algorithm is expressed as follows:

[0048] ,

[0049] in, This represents the total power that can be transferred from the i-th weak section. Let i be the set of connecting lines adjacent to the i-th cross section. These are the voltage amplitudes of the busbars on both sides of the cross-section. For the line The reactance value, These are the phase angles of the bus voltages on both sides of the cross-section;

[0050] ,

[0051] in, This is the cumulative amount of supply imbalance during all time periods. The duration of the scheduling cycle. This represents the total number of weak cross sections. The transmission power limit for the i-th section, For the integration time variable; by minimizing Determine the optimal power transfer ratio for each section.

[0052] Specifically, the first formula in the cross-section power transfer coordination and balancing algorithm is used to calculate the total power that can be transferred from a single weak section. It calculates the total power by multiplying the voltage amplitudes of the busbars on both sides of the section by the line reactance as a base coefficient, then multiplying by the sinusoidal value of the phase angle difference between the two sides, and summing the results for all tie lines belonging to that section. The physical essence of this formula is based on the active power transmission equation of AC transmission lines, and its form originates from the power flow relationship in power system flow calculations. This is because the active power that each tie line can transmit is directly proportional to the voltage product and inversely proportional to the line reactance, while also being modulated by the sinusoidal value of the phase angle difference. Superimposing the transmission capacity of all lines in the same section yields the overall transfer capacity that the section can provide. The second formula is used to quantify the degree of power transfer imbalance among multiple sections throughout the entire time period. For each weak section, the deviation between its total transferable power and the transmission power limit is taken, divided by the limit, and the square is calculated. The squared deviation values ​​of all sections are then integrated over time along the scheduling cycle. In this formula, summation and integration occur in two separate formulas, with the integration time step set to the actual sampling interval of 100 milliseconds. During implementation, the voltage amplitude and phase angle measurements of the busbars on both sides of each section, as well as the line reactance parameters, are first read from the energy management system and substituted into the first formula to calculate the total transferable power for each section. After substituting the total transferable power of each section with the limit value into the second formula to calculate the accumulated imbalance, a particle swarm optimization algorithm is used to search for the optimal allocation coefficient of the transferred power for each section within the range of 0.6 to 1.4 times the limit. The inertia weight is set to 0.8, the acceleration constant is set to 1.5, and the population size is set to 50 particles. After iterative convergence, the operation sequence is generated by arranging the allocation coefficients in descending order.

[0053] Preferably, the dynamic balance model for load redistribution and supply-demand equilibrium is expressed as follows:

[0054] ,

[0055] in, To account for the net power difference of the isolated grid after the adjustment, As the reference voltage, Operating voltage This is the partial derivative of the load power with respect to the voltage. This represents the total number of adjustable generating units. This is the output adjustment amount for the k-th unit. To reduce the total number of load nodes, Let m be the original load reduction amount for the m-th load node. This is the load frequency regulation coefficient. This is the frequency deviation. For the rated frequency; make The allocation scheme that approaches zero is the solution for dynamic supply and demand balance.

[0056] Specifically, the dynamic balance model for load redistribution and supply-demand equilibrium establishes a quantitative expression for the net power difference in isolated grids after redistribution. The formula consists of three core components combined through integration and summation operations. The first part is the load voltage sensitivity term, which performs integration and product operations on the partial derivative of load power with respect to voltage within the range from the reference voltage to the operating voltage. This describes the natural fluctuation of load power caused by voltage changes. The specific value of this partial derivative is determined according to the load type: 2.0 for constant impedance loads, 1.0 for constant current loads, and 0.0 for constant power loads. In the actual system, the comprehensive partial derivative value is obtained by weighting according to the component proportion of each load node. The second part is the total output adjustment of the sending-end units. The output adjustment amounts of all adjustable units are summed and accumulated. The feasible upper limit of the adjustment amount of each unit is determined by the difference between the current output and the maximum technical output of the unit, where the maximum technical output is 95% of the rated capacity. The third part is a correction factor, which is the sum of the total load reduction multiplied by the load frequency adjustment coefficient and the ratio of the frequency deviation to the rated frequency. This correction factor reflects the reduction effect of the actual load reduction when the frequency decreases. The typical value of the load frequency adjustment coefficient is between 0.5 and 2.0, and the frequency deviation is taken from the difference between the real-time frequency and the rated frequency of 50 Hz. Substituting these three parts into the formula and letting the net difference approach zero is the dynamic equilibrium condition. During implementation, the solution is obtained iteratively in the dispatch strategy optimization module. The load reduction and the unit increase are treated as undetermined variables. The trust region reflection algorithm is used to search for the variable combination that minimizes the absolute value of the net difference within the variable boundary constraints. In the boundary constraints, the upper limit of the load reduction is 30% of the current load value, and the upper limit of the unit increase is 10% of the rated capacity. After the iteration converges, the coordinated dispatch scheme is output.

[0057] Preferably, the islanded network transient stability dual-layer constraint calculation platform includes a hybrid architecture computing unit composed of a digital signal processor and a field-programmable gate array. In the upper-layer frequency and voltage dynamic constraint verification, the frequency deviation iteration convergence threshold is set to 0.02, and the voltage deviation iteration convergence threshold is set to 0.01. In the lower-layer power angle stability constraint verification, the maximum allowable deviation of the relative power angle swing curve is set to 0.15. The global convergence judgment condition of the dual-layer iteration is that the weighted sum of the squares of the upper-layer verification residual and the lower-layer verification residual is less than 0.005. The islanded network transient stability dual-layer constraint calculation platform reads the allocation scheme parameters cyclically at a 1-millisecond cycle and outputs the feasible control strategy identification code marked by the dual-layer verification.

[0058] Preferred, such as Figure 2As shown, S2 includes the following sub-steps: S21, extracting the active power output value of each generator and the load power value of each node from the state feature set, and calculating the instantaneous power deviation sequence of the isolated network; S22, using the power deviation sequence as input, approximating the cumulative entropy value of power imbalance using the trapezoidal integral method; S23, generating an instability risk factor based on the ratio of the cumulative entropy value to the preset entropy threshold value, combined with the power deviation time change rate; S24, marking the time when the instability risk factor is greater than 1 as the risk period, and simultaneously locating the cross sections where the transmission power exceeds the limit within the risk period as the set of weak cross sections.

[0059] Specifically, S2 includes the following sub-steps: S21 Extracts the active power output value of each generator and the load power value of each node from the state feature set, and calculates the instantaneous power deviation sequence of the isolated network. This step continuously reads the real-time values ​​of the active power output measurement points of generators and the power measurement points of load nodes with a sampling period of 20 milliseconds. The active power output values ​​of all generators at the same time are added to obtain the total power generation, and the power values ​​of all load nodes are added to obtain the total load power. The instantaneous power deviation value at that time is obtained by subtracting the total load power from the total power generation. The deviation values ​​of all sampling times are arranged sequentially along the time axis to form a deviation sequence. S22 Using the power deviation sequence as input, the trapezoidal integral method is used to approximate the cumulative entropy value of power imbalance. The ratio of the deviation value of two adjacent sampling points to the rated power reference value is multiplied by the natural logarithm of the ratio and the average value is taken. Then, it is multiplied by the sampling interval time of 20 milliseconds. The calculation results of all adjacent point pairs are accumulated along the time axis to obtain the approximate result of the cumulative entropy value. S23 generates an instability risk factor based on the ratio of the accumulated entropy value to a preset entropy threshold, combined with the power deviation time change rate. The entropy threshold is set to 1.2. The power deviation time change rate is calculated by dividing the difference between the deviation value at the current moment and the deviation value at the previous sampling moment by the sampling interval. This change rate is multiplied by an attenuation coefficient of 0.3, and the exponential function value is obtained. The risk factor is obtained by multiplying the entropy ratio by this exponential function value. S24 marks the moments when the instability risk factor is greater than 1 as risk periods, and simultaneously locates the sections with transmission power exceeding the limit within the risk period as a set of weak sections. The risk factor value is judged moment by moment, and time intervals with consecutive values ​​greater than 1 are merged and marked as risk periods. Within this period, the actual transmission power value of all sections is scanned one by one. Sections whose actual values ​​exceed 80% of the rated transmission power limit of the section are included in the set of weak sections and their section numbers are recorded.

[0060] Preferred, such as Figure 3As shown, S3 includes the following sub-steps: S31, for each section in the set of weak sections, read the voltage amplitude and phase angle data of the busbars at both ends of its tie line, and calculate the single-line transferable power value of each tie line; S32, collect the transferable power values ​​of the tie lines belonging to the same section to generate the total transferable power of the section; S33, calculate the cumulative amount of power transfer imbalance throughout the time period based on the deviation of the total transferable power of each section from the transmission power limit; S34, with minimizing the cumulative amount of imbalance as the optimization objective, solve the optimal allocation coefficient of the power transfer of each section by gradient descent method, and generate the power transfer operation sequence by arranging the coefficients in descending order.

[0061] Specifically, S3 includes the following sub-steps: S31 For each section in the set of weak sections, read the voltage amplitude and phase angle data of the busbars at both ends of its tie line, calculate the single-line transferable power value of each tie line, obtain the per-unit value data of the voltage amplitude from the wide-area measurement system, retaining 4 decimal places, and the degree data of the voltage phase angle, and at the same time read the line reactance parameter, divide the product of the voltage amplitudes on both sides by the reactance value, and then multiply by the sine function value of the phase angle difference on both sides to obtain the single-line transmission power value. S32 Collect the transferable power values ​​of the tie lines belonging to the same section to generate the total transferable power of the section, and sum the single-line transmission power values ​​of all tie lines belonging to the same section number to obtain the total transferable power value that the section can provide as a whole. S33 calculates the cumulative power imbalance over the entire time period based on the deviation between the total power transferable at each section and the transmission power limit. For each weak section, the difference between its total power transferable and transmission power limit is divided by the limit and squared. The squared values ​​of all sections are accumulated over the scheduling cycle with a time step of 100 milliseconds to obtain the cumulative amount. S34 takes minimizing the cumulative imbalance as the optimization objective and solves for the optimal allocation coefficient of the power transfer for each section using the gradient descent method. The transfer operation sequence is generated by arranging the coefficients in descending order. The iteration step size is 0.01, and the maximum number of iterations is 500. After convergence, each section obtains an allocation coefficient in the range of 0 to 1. All sections are sorted in descending order of coefficients, and the tie line with the largest power transferable for each section is matched as the preferred channel. The section number and channel number are combined into a sequence.

[0062] Preferred, such as Figure 4As shown, S4 includes the following sub-steps: S41, read the voltage operating value and reference value of the receiving-end node involved in the power transfer operation sequence, and perform voltage-sensitive reduction pre-calculation on the receiving-end load power; S42, read the current upper and lower limits of the output of the adjustable units at the sending end, and determine the feasible range of the output increase of each unit; S43, substitute the load reduction and unit increase into the supply and demand dynamic balance equation, and calculate the net difference of the isolated grid power after the allocation; S44, use the Newton-Raphson method to iteratively adjust the numerical ratio of the reduction and increase until the absolute value of the net difference is lower than the convergence threshold, and output the current ratio as the coordinated allocation scheme.

[0063] Specifically, S4 includes the following sub-steps: S41 Reads the voltage operating value and reference value of the receiving-end node involved in the power transfer operation sequence, performs voltage-sensitive reduction pre-calculation on the receiving-end load power, extracts the current per-unit voltage value and reference voltage value of the bus where the receiving-end node is located from the sequence, reads the current power value and load voltage characteristic parameters of the load connected to the bus, and calculates the natural fluctuation of load power caused by voltage change within the range from the reference voltage to the current voltage as the estimated reduction amount. S42 Reads the current upper and lower limits of the output of the adjustable units at the sending end, determines the feasible range of the output increase of each unit, obtains the maximum technical output value and the current actual output value of each adjustable unit from the unit control system, and uses the difference between the maximum technical output value and the current actual output value as the upper limit of the output increase of the unit, with the lower limit uniformly set to 0. S43 substitutes the load reduction and unit increase into the supply-demand dynamic balance equation to calculate the net power difference of the isolated grid after allocation. The total reduction is obtained by summing the reductions of all receiving-end nodes, and the total increase is obtained by summing the increases of all units. The net difference is obtained by subtracting the total reduction, voltage sensitivity term, and frequency correction term from the total increase. S44 uses the Newton-Raphson method to iteratively adjust the ratio of reductions to increases until the absolute value of the net difference is below the convergence threshold. The current ratio is then output as the coordinated allocation scheme. The convergence threshold is 0.001. Each iteration adjusts the allocation weights of reductions and increases in 0.01 steps according to the positive or negative direction of the net difference. After reaching the convergence condition, the reduction of each node and the increase of each unit are recorded.

[0064] Preferred, such as Figure 5As shown, S5 includes the following sub-steps: S51, mapping the coordinated allocation scheme to frequency response curves and voltage response curves, and the upper-level calculation platform performs dynamic constraint verification of frequency deviation peak value and voltage drop depth; S52, if the upper-level verification passes, the scheme is transferred to the lower-level calculation platform, and the lower-level platform performs stability constraint verification of the generator power angle relative swing curve; S53, if the lower-level verification fails, the power angle instability indicator is fed back to the upper-level platform, and the upper-level platform corrects the power allocation weight in the allocation scheme accordingly; S54, repeating the iterative correction process between the upper and lower levels until the allocation scheme simultaneously meets the frequency and voltage dynamic constraints and the power angle stability constraints, and marking the scheme as a feasible control strategy.

[0065] Specifically, S5 includes the following steps: S51 Mapping the coordinated dispatch scheme to frequency response curves and voltage response curves, the upper-level calculation platform performs dynamic constraint verification of frequency deviation peak value and voltage drop depth, uses the generator speed governor transfer function model and load frequency characteristic model for joint simulation to generate frequency response curves, uses the excitation regulator transfer function model and load voltage characteristic model for joint simulation to generate voltage response curves, and extracts the maximum absolute value of frequency deviation and the maximum relative value of voltage drop depth from the curves. S52 If the upper-level verification passes, the scheme is transferred to the lower-level calculation platform, the lower-level platform performs stability constraint verification of the generator power angle relative swing curve, uses time-domain simulation method to calculate the deviation value of each generator rotor electrical angle relative to the reference machine at each time point, plots the swing curve over the entire simulation time and extracts the maximum deviation value. S53 If the lower-level verification fails, the power angle instability indicator is fed back to the upper-level platform, the upper-level platform corrects the power allocation weight in the dispatch scheme accordingly, shifts the power transfer allocation weight in the original scheme to the section farther from the unstable unit by 0.05 and generates a corrected scheme to be resubmitted to the upper-level verification. S54 repeats the iterative correction process between the upper and lower layers until the allocation scheme simultaneously satisfies the frequency and voltage dynamic constraints and the power angle stability constraints. The scheme is then marked as a feasible control strategy. Each complete iteration cycle takes no more than 200 milliseconds. The peak frequency deviation threshold is 0.5 Hz, the voltage drop depth threshold is 0.1 per unit, and the maximum power angle deviation threshold is 15 degrees. After all constraints are met, a feasible strategy identifier code is output.

[0066] A method for coordinated control of power grid event stability based on load transfer between different locations is implemented through a coordinated control system for power grid event stability based on load transfer between different locations. This method includes: a power grid operation status feature acquisition unit, whose input is connected to the power grid dispatch data network to obtain topology connection status, generator output range, bus voltage phasor, and cross-sectional power limit; and its output is a set of status features. A power grid instability risk identification unit, whose input is connected to the output of the feature acquisition unit, internally embedding a power instability critical entropy criterion algorithm to output risk period identifiers and a set of weak cross-sections. A cross-section transfer strategy generation unit, whose input is connected to the output of the risk identification unit, internally embedding a cross-section transfer power collaborative balancing algorithm to generate a transfer operation sequence. The system outputs to the load dispatching unit; the load off-site dispatching supply and demand balance solution unit, whose inputs are connected to the output of the transfer strategy generation unit and the load parameter port of the dispatching data network, internally embedding the load off-site dispatching supply and demand dynamic balance model to output a coordinated dispatching scheme; the transient stability dual-layer constraint verification unit, whose input is connected to the output of the supply and demand balance solution unit, internally integrating a frequency and voltage dynamic constraint verification module and a power angle stability constraint verification module, outputting a feasible control strategy after dual-layer iterative calculation; and the pre-control and stability command output unit, whose input is connected to the output of the verification unit, and whose output is connected to the automatic generation control device and the stability control device, breaking down the feasible control strategy into pre-control parameters and post-stability action parameters and issuing them separately.

[0067] A collaborative control method for isolated grid events based on load relocation in power grids establishes a complete technical chain from proactive risk identification to control strategy generation and dual-constraint verification. This transforms isolated grid control from passively responding to faults to proactively predicting risks and adjusting operating conditions in advance, significantly improving the predictability and systematic nature of control. The method combines inter-section power coordinated balancing with load relocation, fully leveraging the multi-path relocation potential within the grid interconnection structure. From a global perspective, it achieves optimal matching between generation-side output adjustments and receiving-end load reductions, thus maintaining isolated grid power balance with minimal control costs. Through a transient stability dual-layer constraint accounting mechanism, the method performs dual verification of strategy effectiveness—both dynamic frequency and voltage and static power angle—before the actual issuance of control commands, eliminating any relocation schemes that do not meet stability boundaries and ensuring the reliability and safety of control measures from the outset.

[0068] This invention addresses the problems of offline contingency plans failing to cover the continuously changing fault space and the lack of coordinated consideration in control measures. It uses real-time operational status characteristics as input to drive an entropy criterion algorithm to dynamically identify risks, enabling the control strategy to adaptively adjust according to actual operating conditions. Simultaneously, a cross-section balancing algorithm incorporates the transfer capacity of each connecting line into a unified optimization framework, effectively avoiding local section overload and mismatch of control resources. To address the issues of local control relying on a single threshold and the lack of global coordination among devices leading to over- or under-adjustment, this invention distributes pre-control commands and post-stability control commands in a time-series layer. Pre-control proactively increases the system's stability margin, reducing the probability and amount of post-action actions. Post-stability control intervenes only as a last line of defense when pre-control cannot completely suppress large disturbances. Both achieve parameter-level coordination through feasible strategies output by the same dual-layer constraint verification platform, fundamentally eliminating the uncertainty in control effects caused by multiple devices operating independently.

[0069] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0070] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for coordinated control of power grid event stability based on load transfer to different locations, characterized in that, Includes the following steps: S1, collects topology connection status, generator active power output range, bus voltage amplitude phase angle and cross-sectional transmission power limit from the grid operation and dispatch layer to construct an islanded grid operation status feature set; S2, based on the critical entropy criterion algorithm for power instability of isolated grids, performs transient stability margin quantitative assessment on the state feature set, and identifies the risk period and set of weak sections that may induce power fluctuations to cause isolated grid collapse; S3, for the set of weak cross sections, start the cross-section power transfer collaborative equalization algorithm, calculate the transferable power amplitude and transfer path priority of each tie line, and generate the power transfer operation sequence; S4, call the load off-site allocation supply and demand dynamic balance model, and use the transferred power as input to solve the coordinated allocation scheme of the load reduction at the receiving end and the output increase of the sending end unit. S5 embeds the allocation scheme into the islanded network transient stability dual-layer constraint calculation platform. The upper layer performs frequency and voltage dynamic constraint verification, and the lower layer performs power angle stability constraint verification. Through dual-layer iteration, feasible control strategies that satisfy all constraints are obtained. S6 issues pre-control commands and post-stabilization control commands to the automatic device based on feasible control strategies to carry out full-process control of the isolated grid operation status.

2. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, The algorithm for determining the critical entropy criterion for power instability in isolated networks is expressed as follows: in, The cumulative entropy value for power imbalance. This represents the start time of the disturbance. This is the critical resection time. Let be the instantaneous deviation between the total generating power and the total load power within the isolated grid at time t. This is the reference value for the rated power of an isolated grid; in, As an instability risk factor, To preset the entropy threshold, The attenuation coefficient is... The time rate of change of power deviation at the critical cutoff moment; when If the value is greater than 1, the current operating state is considered to have a risk of instability.

3. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, The cross-section power transfer collaborative equalization algorithm is expressed as follows: in, This represents the total power that can be transferred from the i-th weak section. Let i be the set of connecting lines adjacent to the i-th cross section. These are the voltage amplitudes of the busbars on both sides of the cross-section. For the line The reactance value, These are the phase angles of the bus voltages on both sides of the cross-section; in, This is the cumulative amount of supply imbalance during all time periods. The duration of the scheduling cycle. This represents the total number of weak cross sections. The transmission power limit for the i-th section, For the integration time variable; by minimizing Determine the optimal power transfer ratio for each section.

4. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, The dynamic balance model for load redistribution and supply-demand equilibrium is expressed as follows: in, To account for the net power difference of the isolated grid after the adjustment, As the reference voltage, Operating voltage, This is the partial derivative of the load power with respect to the voltage. This represents the total number of adjustable generating units. This is the output adjustment amount for the k-th unit. To reduce the total number of load nodes, Let m be the original load reduction amount for the m-th load node. This is the load frequency regulation coefficient. This is the frequency deviation. For the rated frequency; make The allocation scheme that approaches zero is the solution for dynamic supply and demand balance.

5. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, The isolated network transient stability dual-layer constraint calculation platform includes a hybrid architecture computing unit composed of a digital signal processor and a field-programmable gate array. In the upper-layer frequency and voltage dynamic constraint verification, the frequency deviation iteration convergence threshold is set to 0.02, and the voltage deviation iteration convergence threshold is set to 0.

01. In the lower-layer power angle stability constraint verification, the maximum allowable deviation of the relative power angle swing curve is set to 0.

15. The global convergence judgment condition of the dual-layer iteration is that the weighted sum of the squares of the upper-layer verification residual and the lower-layer verification residual is less than 0.

005. The isolated network transient stability dual-layer constraint calculation platform reads the allocation scheme parameters cyclically at a 1-millisecond cycle and outputs the feasible control strategy identification code marked by the dual-layer verification.

6. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, S2 includes: The active power output of each generator and the load power of each node are extracted from the state feature set to calculate the instantaneous power deviation sequence of the isolated network; the cumulative entropy value of power imbalance is approximated by the trapezoidal integral method using the power deviation sequence as input. An instability risk factor is generated based on the ratio of the cumulative entropy value to a preset entropy threshold, combined with the power deviation time change rate. The time when the instability risk factor is greater than 1 is marked as the risk period, and the cross sections where the transmission power exceeds the limit during the risk period are simultaneously located as the set of weak cross sections.

7. The method for coordinated control of power grid isolated events based on load transfer in accordance with claim 1, characterized in that, S3 includes the following steps: For each section in the set of weak sections, read the voltage amplitude and phase angle data of the busbars at both ends of its tie line, and calculate the single-line transferable power value of each tie line; collect the transferable power values ​​of the tie lines belonging to the same section to generate the total transferable power of that section; Based on the degree of deviation between the total transferable power at each section and the transmission power limit, calculate the cumulative amount of power imbalance throughout the entire time period; With minimizing the cumulative imbalance as the optimization objective, the optimal allocation coefficient of the power transfer at each section is solved by the gradient descent method, and the power transfer operation sequence is generated by arranging the coefficients in descending order.

8. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, S4 includes the following steps: Read the voltage operating value and reference value of the receiving node involved in the power transfer operation sequence, and perform voltage-sensitive reduction pre-calculation on the receiving load power; Read the current upper and lower limits of the adjustable units at the sending end to determine the feasible range of the output increase of each unit; substitute the load reduction and the unit increase into the supply and demand dynamic balance equation to calculate the net power difference of the isolated grid after the allocation; use the Newton-Raphson method to iteratively adjust the numerical ratio of the reduction and increase until the absolute value of the net difference is lower than the convergence threshold, and output the current ratio as the coordinated allocation scheme.

9. The method for coordinated control of isolated grid events based on load transfer in accordance with claim 1, characterized in that, S5 includes the following steps: The coordinated dispatch scheme is mapped to frequency response curves and voltage response curves. The upper-level calculation platform performs dynamic constraint verification of frequency deviation peak value and voltage drop depth. If the upper-level verification passes, the scheme is transferred to the lower-level calculation platform, where the stability constraint verification of the generator power angle relative swing curve is performed. If the lower-level verification fails, the upper-level layer is fed back with a power angle instability flag. The upper-level layer then corrects the power allocation weights in the allocation scheme accordingly. The iterative correction process between the upper and lower layers is repeated until the allocation scheme simultaneously meets the frequency and voltage dynamic constraints and the power angle stability constraints. The scheme is then marked as a feasible control strategy.

Images