Networking type converter control method, system, device and medium with low communication cost

Abstract

This invention discloses a low-communication-cost control method, system, device, and medium for grid-connected converters. The method includes: constructing a frequency-voltage-power closed-loop control model and pre-setting operating parameter ranges; real-time acquisition of local operating quantities, comparing them with the previous broadcast value, calculating the error norm, and determining whether to trigger a communication event; estimating system characteristic quantities and the upper bound of coupling strength when a communication event occurs; dynamically adjusting the trigger threshold and trigger function parameters to ensure the lower bound of the event interval; superimposing a neighbor node state correction term on the local droop control, updating only when an event is triggered; implementing soft / hard limiting on voltage, frequency, and power; increasing the communication trigger frequency when the operating point approaches the over-limit range, and resuming low communication mode after the over-limit is cleared; continuing low communication mode after disturbance elimination. This invention can provide key data support for grid-connected control of power grids and has significant implications in the field of grid-connected control of power systems.

Description

Technical Field

[0001] This invention belongs to the field of converter control, and specifically relates to a low-communication-cost network converter control method, system, equipment and medium, which is suitable for the coordinated operation of multiple network converters in islanded or weakly interconnected systems. Background Technology

[0002] Grid-based converters are widely used in microgrids and weakly interconnected systems with a high proportion of renewable energy due to their ability to actively establish voltage and frequency references and provide virtual inertia and voltage support. To ensure the coordinated operation of multiple grid-based converters and achieve frequency stability, voltage recovery, and power distribution, control strategies are generally divided into virtual synchronous machine control, droop control, model predictive control, and data-driven control based on reinforcement learning. The first two methods are simple in structure and easy to implement, but the control parameters are fixed and usually rely on high-frequency communication to complete secondary voltage and frequency recovery; while model predictive control and data-driven methods can adapt to uncertainties, they require centralized computing or high-bandwidth real-time data support.

[0003] The stable operation of microgrids and weakly interconnected systems highly depends on the reliability and scalability of the control architecture. Centralized control suffers from single-point failure risks and poor scalability, and has been gradually replaced by distributed control. Distributed control achieves global objectives through information exchange between local controllers, improving system flexibility and fault tolerance. However, traditional distributed methods often employ fixed-period sampling and full-state communication. Frequent data exchange not only consumes a large amount of communication bandwidth but also easily leads to problems such as delays and packet loss, affecting system dynamic performance and increasing network security risks.

[0004] Especially in weak power grids dominated by large-scale GFM (Geostationary Ground Facility), the update frequency of control commands directly affects system stability. Traditional periodic communication strategies suffer from redundant data transmission issues, while existing event-triggered methods still have shortcomings in terms of complete distribution, lack of global parameter dependence, and explicit performance lower bound guarantees. Therefore, there is an urgent need for a grid-type converter control method that combines frequency-voltage stability control, power sharing, dynamic adaptive triggering, and safety protection functions under low communication conditions to meet the stable operation requirements under environments with limited communication, changing grid topology, and disturbances. Summary of the Invention

[0005] To address at least one of the problems existing in the prior art, this invention provides a low-communication-cost grid-type converter control method, system, device, and medium. It is applicable to the collaborative operation of grid-type converters in islanded or weakly interconnected power systems powered by 100% renewable energy. It can provide key data support for grid-type control of the power grid and has significant implications in the field of grid-type control of power systems.

[0006] To achieve the objective of this invention, a low-communication-cost network converter control method is provided, comprising the following steps:

[0007] A low-communication-cost network converter control method includes the following steps:

[0008] S1. Construct a frequency-voltage-power closed-loop control model and preset the operating parameter range;

[0009] S2. Real-time collection of local running volume, comparison with the previous broadcast value, calculation of error norm, and combination with exponential decay threshold to form a trigger function to determine whether to trigger a communication event;

[0010] S3. When a communication event occurs, estimate the system characteristic quantities and the upper bound of the coupling strength based on local measurement results;

[0011] S4. Dynamically adjust the exponential decay threshold and trigger function parameters based on the stability margin to ensure the lower bound of the event interval;

[0012] S5. Based on the local droop control, the neighbor node status correction item is superimposed and updated only when the event is triggered, so as to realize the coordinated recovery of frequency and voltage of multiple machines and power distribution.

[0013] S6. Implement soft / hard limiting on voltage, frequency, and power. When the operating point approaches the over-limit range, increase the communication trigger frequency. After the over-limit is released, return to low communication mode.

[0014] S7. After the disturbance is eliminated, continue in low communication mode until the next trigger condition is met.

[0015] Furthermore, the frequency-voltage-power closed-loop control model includes:

[0016] Based on the system's rated frequency, the frequency command is calculated using the primary frequency droop coefficient according to the active power deviation.

[0017] Based on the system rated voltage, the voltage command is calculated using the voltage droop factor according to the reactive power deviation. ;

[0018] A low-pass filter is used to extract the average value of active power and reactive power, and it is compared with the frequency command and voltage command to obtain the deviation input to the droop control.

[0019] The inner loop controls the instantaneous value of the converter's output current, while the outer loop controls the converter's output voltage.

[0020] Furthermore, step S2 includes the following sub-steps:

[0021] Acquire local operating parameters, including real-time frequency, bus voltage, active power and reactive power, to form a set of operating parameters;

[0022] Save the operating parameters from the last external broadcast and calculate the weighted Euclidean norm error between the current operating quantity and the last broadcast value;

[0023] The exponential decay threshold is calculated based on the initial threshold, the minimum allowable threshold, and the decay coefficient.

[0024] Trigger function formation: If the weighted Euclidean norm error is greater than the exponential decay threshold, it is determined to be an event that requires communication, and the current set of operating parameters is immediately broadcast to neighboring nodes and the trigger time is recorded; if the weighted Euclidean norm error is less than or equal to the exponential decay threshold, it remains silent and does not send data.

[0025] Furthermore, in step S3, when a communication event occurs, a distributed consensus algorithm is used to estimate the system characteristic quantities and the upper bound of coupling strength based on local measurement results, specifically as follows:

[0026] Based on the set of operating parameters and droop control parameters obtained from the current sampling, calculate the local linearized state-space model;

[0027] The maximum real part eigenvalue and condition number are obtained by calculating the closed-loop eigenvalue set of the system matrix locally.

[0028] When an event is triggered, the maximum real part eigenvalue and the condition number are sent to the directly adjacent node via point-to-point communication. The adjacent node simultaneously sends its eigenvalues. After the two sides exchange them, the maximum consistency algorithm is iterated to obtain the maximum real part eigenvalue and the maximum condition number of the entire network.

[0029] Based on the maximum real part eigenvalue and the maximum condition number of the entire network, the upper bound of the coupling strength of the system is obtained.

[0030] Further, in step S4, the grid-type converter receives the maximum real part eigenvalue and the maximum condition number of the entire network calculated in step S3, and combines them with the dynamic correction event triggering criterion for the upper bound of coupling strength, including:

[0031] The stability margin coefficient is defined based on the maximum real part eigenvalue of the entire network and a preset stability security threshold.

[0032] The stability margin coefficient and the decision boundary for sufficient or insufficient margin are compared to adjust the initial value parameter of the exponential decay threshold.

[0033] Furthermore, step S5 includes the following sub-steps:

[0034] Using the frequency-voltage-power closed-loop control model established in step S1, calculate the frequency command and voltage command respectively;

[0035] Correction terms for voltage and frequency are constructed based on the state differences between adjacent nodes;

[0036] Active power and reactive power are allocated proportionally respectively;

[0037] Overlay the corrections onto the local control baseline:

[0038]

[0039] in, , To incorporate the frequency and voltage control commands with distributed correction, , These are the correction gain coefficients for frequency and voltage, respectively. For frequency commands, This is a voltage command.

[0040] Furthermore, step S7 includes:

[0041] Continuously monitor local workload and the status information of neighboring nodes. If all workloads are within a safe range and the deviation from the reference value is less than the static deviation threshold, it is determined that the disturbance has been eliminated.

[0042] Once the disturbance is eliminated, the initial threshold for event triggering is restored to the set normal threshold value, the threshold attenuation coefficient is set to a low value, the triggering interval is extended, and communication is only performed when the triggering function of step S2 is triggered.

[0043] If no event is triggered within a preset time, a health status packet is periodically sent to neighboring nodes.

[0044] The present invention provides a low-communication-cost network-type converter control system, comprising the following modules:

[0045] The local measurement and execution module is used to collect voltage and current signals at the output of the grid-type converter in real time, calculate instantaneous active and reactive power, detect local operating frequency and bus voltage, input the collected operating quantities into the local execution unit, and drive the power conversion unit to perform regulation according to control commands.

[0046] The event-triggered communication module is used to calculate the deviation norm between the current running volume and the previous broadcast value, generate an exponential decay threshold value, determine whether the threshold is exceeded and trigger a communication event, and send the current running data to the neighboring node through a point-to-point communication link;

[0047] The adaptive parameter estimation module is used to calculate the linearized system matrix based on the local operating state and obtain the maximum real part eigenvalue and condition number. It also exchanges information with neighboring nodes to obtain the maximum eigenvalue and condition number of the entire network and estimates the upper bound of the system coupling strength.

[0048] The trigger parameter update module is used to determine whether the system state needs to increase or decrease the trigger frequency based on the stability margin coefficient, dynamically adjust the initial threshold value and exponential decay coefficient, and set and execute the minimum time interval of events.

[0049] The distributed coordination control module is used to calculate frequency and voltage commands based on local power deviation, receive the operating status of neighboring nodes, calculate frequency and voltage correction terms, allocate active and reactive power reference values ​​according to power ratio, and output the corrected control commands to the local execution module to realize multi-machine collaborative operation.

[0050] The protection and limiting module is used to monitor frequency, voltage, active power, and reactive power in real time and compare them with preset limits. When the limit is approaching, a soft limiting strategy is executed. When the limit is exceeded, a hard limiting or emergency shutdown strategy is executed. During the over-limit period, the communication trigger frequency is increased, and after the limit is lifted, the low communication mode is restored.

[0051] The present invention provides an apparatus comprising a processor and a memory, the memory for storing instructions or computer programs, and the processor for executing the instructions or computer programs in the memory to cause the apparatus to perform the steps of the method.

[0052] The present invention provides a computer-readable storage medium storing instructions that, when executed on a device, cause the device to perform the steps of the method.

[0053] The present invention has the following advantages over the prior art:

[0054] (1) The present invention uses an event triggering mechanism based on exponential decay threshold and exponential decay to broadcast data only when the operating state changes significantly, thereby avoiding redundant information exchange in traditional periodic communication and significantly reducing communication load.

[0055] (2) The present invention uses local computing and distributed consensus algorithm to realize the stability margin estimation of the whole network. It does not rely on centralized information such as the eigenvalues ​​of the global communication topology matrix, and can maintain the scalability and robustness of the control architecture when the system topology changes, large-scale node access or node exit. Attached Figure Description

[0056] Figure 1 A flowchart of a low-communication-cost network converter control method provided in an embodiment of the present invention.

[0057] Figure 2 This is a schematic diagram of a low-communication-cost network converter control system provided in an embodiment of the present invention.

[0058] Figure 3This is a schematic diagram of an islanded power grid powered by 100% renewable energy in an embodiment of the present invention.

[0059] Figure 4 A comparison chart of control results from different methods. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Please see Figure 1 The present invention provides a low-communication-cost network converter control method, comprising the following steps:

[0062] S1) Establish a frequency-voltage-power closed-loop control model within each converter and preset the operating parameter range.

[0063] The operating parameter range includes upper and lower limits for frequency, upper and lower limits for voltage, and power limits.

[0064] This step establishes a closed-loop control framework, including droop control, power control, voltage regulation, and frequency regulation, within the local controller of each grid-connected converter. This closed-loop control framework is constructed through the following sub-steps:

[0065] S11, at the system rated frequency Based on the active power deviation, the primary frequency droop coefficient is used as a benchmark. Frequency calculation instructions , Active power This is a reference value for the target active power.

[0066] S12, based on the system rated voltage Based on the reactive power deviation, the voltage droop coefficient is used as a reference. Calculate voltage command , Reactive power This is a reference value for reactive power;

[0067] S13. Use a low-pass filter to extract active power. With reactive power The average value, and compared with the reference values ​​of frequency and voltage, i.e., the frequency command. Voltage command The deviation is compared and input to the droop control.

[0068] S14. The instantaneous value of the converter output current is controlled by the inner loop, and the output voltage of the converter is controlled by the outer loop to ensure dynamic response and steady-state accuracy.

[0069] S2) Event Trigger Monitoring: Real-time collection of local operating quantities of grid-type converter nodes, comparison with the previous broadcast value, calculation of error norm, and formation of trigger function in combination with exponential decay threshold. In the trigger function, a communication event is triggered when the error exceeds the threshold, and remains silent when it does not exceed the threshold.

[0070] In one embodiment of the present invention, an event-triggered communication determination module is provided in the local controller of each grid-connected converter. The event-triggered communication determination module is configured to perform the following steps:

[0071] S21. Local operating parameter acquisition, real-time frequency. Bus voltage Active power and reactive power , forming nodes Set of running parameters The sampling period is no less than 50 ms;

[0072] S22. Save the set of operating parameters from the last external broadcast. Calculate the weighted Euclidean norm error between the current running volume and the last broadcast value. :

[0073]

[0074] in, , , , These are the weighting coefficients for frequency, voltage, active power, and reactive power, respectively. , , , The first Frequency, voltage, active power, reactive power at any given moment , , , These are the frequency, voltage, active power, and reactive power of the last external broadcast;

[0075] S23. Calculation of Exponential Decay Threshold Value :

[0076] The initial threshold is triggered based on a preset event. Minimum allowed threshold value and attenuation coefficient (Control the threshold descent speed), combined with the time interval since the last triggered event. Calculate the exponential decay threshold value at the current time. Its calculation formula is

[0077] ;

[0078] S24. Forming the trigger function: The function calculated in step S22... Compared with the current exponential decay threshold value calculated in step S23 Compare the results to determine the trigger function;

[0079] like This is then determined to be an event requiring communication, and the current set of operating parameters is immediately broadcast. Reach neighboring nodes and record the trigger time; if Then remain silent and do not send data.

[0080] S3) Stability Margin Estimation: When a communication event occurs, a distributed consensus algorithm is used to estimate the system's characteristic quantities and the upper bound of coupling strength based on local measurement results. Specifically:

[0081] S31. Based on the set of operating parameters obtained from the current sampling. And droop control parameters, calculate the local linearized state-space model :

[0082]

[0083] in, This is the local calculation system matrix (including frequency, voltage, and power dynamic equation coefficients). For the input matrix, To control the input vector;

[0084] S32. Using the Jacobi method, the system matrix is ​​calculated locally. The closed-loop eigenvalue set is used to obtain local feature quantities, including the eigenvalue with the largest real part. and condition numbers ;

[0085] S33. When the event is triggered, the local feature quantity—the eigenvalue with the largest real part—is... and condition numbers The feature values ​​are sent to the directly adjacent nodes via point-to-point communication. The directly adjacent nodes simultaneously send their feature values, and after the two sides exchange them, the maximum consistency algorithm is iterated.

[0086]

[0087] in, Indexing neighboring nodes, For the number of iteration rounds; if the communication topology is a multi-hop network, after a finite number of iterations, each node obtains the eigenvalue with the largest real part of the entire network. With the maximum condition number ; , The first Round iteration node The eigenvalue with the largest real part and the condition number. , The first Round iteration node The eigenvalue of the largest real part and the condition number;

[0088] S34. Calculate the upper bound of the system coupling strength using the stability criterion formula based on the characteristic of the largest real part of the entire network. :

[0089]

[0090] S4) Adaptive update of trigger parameters: The exponential decay threshold and trigger function parameters are dynamically adjusted based on the stability margin to ensure the lower bound of the event interval. Specifically, in the trigger parameter update module of each grid-type converter, the maximum real part eigenvalue and the maximum condition number of the entire network calculated in step S3 are received, and the event triggering criterion is dynamically corrected in combination with the upper bound of the coupling strength. The steps are as follows:

[0091] S41. Set the normal threshold value When the system is running normally and stably, the initial threshold value for event triggering is... Take the standard threshold value , serving as the baseline for triggering parameters;

[0092] In one embodiment of the present invention, a conventional threshold value The value ranges from 0.9 pu to 1.1 pu.

[0093] S42. Define the stability margin coefficient. :

[0094] ;

[0095] in: This is a preset stability and safety threshold; This indicates that the system is stable. This indicates that the system is in a critical or unstable state;

[0096] S43, Threshold adjustment, if Then increase the initial threshold value for event triggering. ;like Then reduce the initial threshold value for event triggering. :

[0097] ;

[0098] in, and This serves as the boundary for determining whether the margin is sufficient or insufficient. and This refers to the percentage adjustment above and below the threshold.

[0099] S5) Distributed Coordination Control: Based on local vertical control, neighbor node state correction items are superimposed and updated only when an event is triggered, realizing coordinated recovery of frequency and voltage and power distribution among multiple machines, specifically:

[0100] S51. Using the frequency-voltage-power closed-loop control model established in step S1), calculate the local frequency reference value and voltage reference value, i.e., the frequency command. Voltage command :

[0101] S52. Construct voltage and frequency correction terms based on the state difference between adjacent nodes:

[0102]

[0103] in, , For nodes Correction frequency and voltage, , For nodes The frequency and voltage, , Adjacent nodes and Coupling weights, The total number of nodes;

[0104] S53. Based on the frequency-active power droop relationship, and according to the primary frequency droop coefficient... Active power is allocated proportionally; based on the voltage-reactive power droop relationship and the voltage droop coefficient... Distribute reactive power proportionally.

[0105] S54. Add the correction to the local control baseline:

[0106]

[0107] in, , To incorporate the frequency and voltage control commands with distributed correction, , These are the corrected gain coefficients for frequency and voltage, respectively.

[0108] S6) Protection and Limiting Execution: Soft / hard limiting is applied to voltage, frequency, and power. When the operating point approaches the over-limit range, the communication trigger frequency is increased. After the over-limit is released, the low communication mode is restored (i.e., closed-loop control is achieved through the frequency-voltage-power closed-loop control model in step S1).

[0109] In one embodiment of the present invention, soft limiting: when the system voltage, frequency, and power exceed the set safety limit by 0.97 times, the change of the corresponding parameters is gradually slowed down; hard limiting: when the system voltage, frequency, and power exceed the set safety limit, the coordinated control strategy of steps S2) to S6) is immediately executed.

[0110] S7) Cyclic Operation: After the disturbance is eliminated, the low communication mode continues until the next trigger condition is met, specifically:

[0111] S71. Continuously monitor local operation volume and the status information of neighboring nodes. If all operation volume is within a safe range and the deviation from the reference value is lower than the static deviation threshold, it is determined that the disturbance has been eliminated.

[0112] S72. After the disturbance is eliminated, restore the initial threshold of the event trigger to the normal threshold value set in S4), set the threshold attenuation coefficient to a low value (such as between 0.01 and 0.05), extend the trigger interval, and only communicate when the trigger condition in S2 is met.

[0113] S73. In the event that no event is triggered for a long time, periodically (e.g., every 5 to 10 seconds) send a health status packet to the neighboring node, which includes: frequency, voltage, active power, reactive power, and power supply status.

[0114] In one embodiment of the present invention, a low-communication-cost network-type converter control system is provided for implementing the method provided in the foregoing embodiments. The system includes the following modules:

[0115] The local measurement and execution module includes a voltage sensor for acquiring voltage, a current sensor for acquiring current, a frequency detection unit for detecting frequency, a power detection unit for detecting power, and a local execution unit for executing local control (executes the content of step S1). It integrates an analog-to-digital converter and a digital signal processor, acquires the three-phase voltage and current signals at the output of the grid-type converter in real time, calculates the instantaneous active power and reactive power, detects the local operating frequency and bus voltage, inputs the acquired operating quantities into the local execution unit, and drives the power conversion unit to perform adjustment according to the control command.

[0116] The event-triggered communication module includes an error calculation unit, a threshold generation unit, a communication interface, and a buffer queue. It is used to calculate the error norm between the current runtime and the previous broadcast value, and to generate an exponentially decaying threshold (threshold value). ), determine whether the threshold has been exceeded and trigger a communication event, and send the current running data to the neighboring node through the point-to-point communication link;

[0117] The adaptive parameter estimation module includes a local state space model construction unit, an eigenvalue solving unit, a condition number calculation unit, and a distributed consensus operation unit. It calculates the linearized system matrix based on the local operating state and obtains the maximum real part eigenvalue and condition number. Through the distributed consensus algorithm, it exchanges information with neighboring nodes to obtain the maximum eigenvalue and condition number of the entire network. Based on the calculation results, it estimates the upper bound of the system coupling strength.

[0118] The trigger parameter update module includes a margin calculation unit, a threshold adjustment unit, and a minimum event interval control unit. It determines whether the system state needs to increase or decrease the trigger frequency based on the stability margin coefficient, dynamically adjusts the initial threshold value, and sets and executes the minimum event interval.

[0119] The distributed coordination control module includes a droop control calculation unit, a neighbor status receiving unit, a distributed correction calculation unit, and a power allocation optimization unit. It calculates frequency and voltage reference values ​​based on local power deviation, receives the operating status of neighbor nodes, calculates frequency and voltage correction terms, allocates active and reactive power reference values ​​according to power ratio, and outputs the corrected control commands to the local execution unit to realize multi-machine collaborative operation.

[0120] The protection and limiting module includes an operation monitoring unit, a soft limiting unit, a hard limiting unit, and a communication strategy switching unit. It monitors frequency, voltage, active power, and reactive power in real time and compares them with preset limits. When the limit is approaching, a soft limiting strategy is executed. When the limit is exceeded, a hard limiting or emergency shutdown strategy is executed. During the over-limit period, the communication trigger frequency is increased, and after the limit is lifted, the low communication strategy is restored.

[0121] The system includes a local measurement and execution module, an event-triggered communication module, an adaptive parameter estimation module, a distributed coordination control module, and a safety protection module. Each converter node forms a point-to-point sparse communication network, eliminating the need for a centralized controller. This invention enables frequency-voltage coordinated control and power distribution of multiple grid-type converters without the need for high-frequency communication, significantly reducing the number of communication events and improving the stability and robustness of the system in communication-constrained scenarios.

[0122] In one embodiment of the present invention, an apparatus is provided, the apparatus including a processor and a memory, the memory for storing instructions or computer programs, and the processor for executing the instructions or computer programs in the memory to cause the apparatus to perform the steps of the method described in the foregoing embodiments.

[0123] In one embodiment of the present invention, a computer-readable storage medium stores instructions that, when executed on a device, cause the device to perform the steps of the method described in the foregoing embodiments.

[0124] In one embodiment of the present invention, as follows: Figure 3 The system shown is analyzed as an example. This system operates in an island-wide mode, comprising three distributed photovoltaic power stations and one distributed wind turbine generator. All devices are connected via a grid-connected power line (GFL) converter, and there is no communication between units. Voltage and frequency reference values ​​are entirely provided by the three grid-connected energy storage (GFM) units, which share a dedicated communication layer, enabling adjacent energy storage units to exchange status information. The system parameters are configured as follows: rated frequency of 50 Hz and rated voltage of 1.0 pu. The proposed event-triggered distributed control strategy is evaluated through comprehensive simulation and compared with traditional control methods (continuous control and centralized control). The results are shown in Table 1 and... Figure 4 As shown.

[0125] Table 1 Comparison of Indicators for Different Methods

[0126]

[0127] As shown in Table 1, the maximum frequency deviation of the method of this invention is 0.8222 Hz, which is significantly better than the 1.0650 Hz of continuous control and the 1.0194 Hz of centralized control. The maximum rotational speed (RoCoF) is only 23.52 Hz / s, which is significantly lower than both continuous control (26.55 Hz / s) and centralized control (34.45 Hz / s), effectively reducing the dynamic impact of the system during disturbances. The maximum power imbalance of the three methods under this operating condition is 1.35 pu, indicating that their performance in terms of power distribution accuracy is comparable.

[0128] In terms of steady-state performance, the frequency control error of the method of this invention is 0.0072 Hz, which is better than the 0.0085 Hz of continuous control and close to the 0.0022 Hz of centralized control. Meanwhile, the method of this invention only experiences 116 communication events throughout the entire operating cycle, a reduction of approximately 92% compared to the 1449 events in continuous control, significantly reducing the communication burden. Compared to centralized control, this invention retains the scalability and resistance to single-point failures of the distributed structure, does not rely on a centralized control center, and possesses higher reliability. The results show that this invention can significantly reduce communication load under communication-constrained conditions while maintaining excellent frequency, voltage stability, and power distribution performance, balancing communication efficiency, dynamic response, and system security.

[0129] The above embodiments verify that the solution provided by the embodiments of the present invention is feasible and effective.

[0130] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A low-communication-cost network converter control method, characterized in that, Includes the following steps: S1. Construct a frequency-voltage-power closed-loop control model and preset the operating parameter range; S2. Real-time collection of local running volume, comparison with the previous broadcast value, calculation of error norm, and combination with exponential decay threshold to form a trigger function to determine whether to trigger a communication event; Includes the following sub-steps: Acquire local operating parameters, including real-time frequency, bus voltage, active power and reactive power, to form a set of operating parameters; Save the operating parameters from the last external broadcast and calculate the weighted Euclidean norm error between the current operating quantity and the last broadcast value; The initial threshold is triggered based on a preset event. Minimum allowed threshold value and attenuation coefficient Combined with the time interval since the last triggered event Calculate the exponential decay threshold value at the current time. Its calculation formula is ； Trigger function formation: If the weighted Euclidean norm error is greater than the exponential decay threshold, it is determined to be an event that requires communication, and the current set of operating parameters is immediately broadcast to neighboring nodes and the trigger time is recorded; If the weighted Euclidean norm error is less than or equal to the exponential decay threshold, then remain silent and do not send data. S3. When a communication event occurs, estimate the system characteristic quantities and the upper bound of the coupling strength based on local measurement results; S4. Dynamically adjust the exponential decay threshold and trigger function parameters based on the stability margin to ensure the lower bound of the event interval; S5. Based on the local droop control, the neighbor node status correction item is superimposed and updated only when the event is triggered, so as to realize the coordinated recovery of frequency and voltage of multiple machines and power distribution. S6. Implement soft / hard limiting on voltage, frequency, and power. When the operating point approaches the over-limit range, increase the communication trigger frequency. After the over-limit is released, return to low communication mode. S7. After the disturbance is eliminated, continue in low communication mode until the next trigger condition is met.

2. The low-communication-cost network converter control method according to claim 1, characterized in that, The frequency-voltage-power closed-loop control model includes: Based on the system's rated frequency, the frequency command is calculated using the primary frequency droop coefficient according to the active power deviation. Based on the system rated voltage, the voltage command is calculated using the voltage droop coefficient according to the reactive power deviation. A low-pass filter is used to extract the average value of active power and reactive power, and it is compared with the frequency command and voltage command to obtain the deviation input to the droop control. The inner loop controls the instantaneous value of the converter's output current, while the outer loop controls the converter's output voltage.

3. The low-communication-cost network converter control method according to claim 1, characterized in that, In step S3, when a communication event occurs, a distributed consensus algorithm is used to estimate the system characteristic quantities and the upper bound of coupling strength based on local measurement results. Specifically: Based on the set of operating parameters and droop control parameters obtained from the current sampling, calculate the local linearized state-space model; The maximum real part eigenvalue and condition number are obtained by calculating the closed-loop eigenvalue set of the system matrix locally. When an event is triggered, the maximum real part eigenvalue and the condition number are sent to the directly adjacent node via point-to-point communication. The adjacent node simultaneously sends its eigenvalues. After the two sides exchange them, the maximum consistency algorithm is iterated to obtain the maximum real part eigenvalue and the maximum condition number of the entire network. Based on the maximum real part eigenvalue and the maximum condition number of the entire network, the upper bound of the coupling strength of the system is obtained.

4. The low-communication-cost network converter control method according to claim 3, characterized in that, In step S4, the grid-type converter receives the maximum real part eigenvalue and maximum condition number of the entire network calculated in step S3, and combines them with the dynamic correction event triggering criterion for the upper bound of coupling strength, including: The stability margin coefficient is defined based on the maximum real part eigenvalue of the entire network and a preset stability security threshold. The stability margin coefficient and the threshold for determining whether the margin is sufficient or insufficient are compared to adjust the exponential decay threshold.

5. A low-communication-cost network converter control method according to claim 1, characterized in that, Step S5 includes the following sub-steps: Using the frequency-voltage-power closed-loop control model established in step S1, calculate the frequency command and voltage command respectively; Correction terms for voltage and frequency are constructed based on the state differences between adjacent nodes; Active power and reactive power are allocated proportionally respectively; Overlay the corrections onto the local control baseline: in, , To incorporate the frequency and voltage control commands with distributed correction, , These are the correction gain coefficients for frequency and voltage, respectively. For frequency commands, This is a voltage command.

6. A low-communication-cost network converter control method according to any one of claims 1-5, characterized in that, Step S7 includes: Continuously monitor local workload and the status information of neighboring nodes. If all workloads are within a safe range and the deviation from the reference value is less than the static deviation threshold, it is determined that the disturbance has been eliminated. Once the disturbance is eliminated, the initial threshold for event triggering is restored to the set normal threshold value, the threshold attenuation coefficient is set to a low value, the triggering interval is extended, and communication is only performed when the triggering function of step S2 is triggered. If no event is triggered within a preset time, a health status packet is periodically sent to neighboring nodes.

7. A low-communication-cost network-type converter control system, characterized in that, For implementing the method according to any one of claims 1-6, the system comprises the following modules: The local measurement and execution module is used to collect voltage and current signals at the output of the grid-type converter in real time, calculate instantaneous active and reactive power, detect local operating frequency and bus voltage, input the collected operating quantities into the local execution unit, and drive the power conversion unit to perform regulation according to control commands. The event-triggered communication module is used to calculate the deviation norm between the current running volume and the previous broadcast value, generate an exponential decay threshold, determine whether the exponential decay threshold is exceeded and trigger a communication event, and send the current running data to the neighboring node through a point-to-point communication link; The adaptive parameter estimation module is used to calculate the linearized system matrix based on the local operating state and obtain the maximum real part eigenvalue and condition number, exchange information with neighboring nodes to obtain the maximum eigenvalue and condition number of the entire network, and estimate the upper bound of the system's coupling strength. The trigger parameter update module is used to determine whether the system state needs to increase or decrease the trigger frequency based on the stability margin coefficient, dynamically adjust the initial threshold value and exponential decay coefficient, and set and execute the minimum time interval of events. The distributed coordination control module is used to calculate frequency and voltage commands based on local power deviation, receive the operating status of neighboring nodes, calculate frequency and voltage correction terms, allocate active and reactive power reference values ​​according to power ratio, and output the corrected control commands to the local measurement and execution module to realize multi-machine collaborative operation. The protection and limiting module is used to monitor frequency, voltage, active power, and reactive power in real time and compare them with preset limits. When the limit is approaching, a soft limiting strategy is executed. When the limit is exceeded, a hard limiting or emergency shutdown strategy is executed. During the over-limit period, the communication trigger frequency is increased, and after the limit is lifted, the low communication mode is restored.

8. A device, characterized in that, The device includes a processor and a memory for storing instructions or computer programs, and the processor for executing the instructions or computer programs in the memory to cause the device to perform the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed on the device, cause the device to perform the steps of the method according to any one of claims 1-6.

Images