Flexible power distribution network distributed collaborative control method and system based on edge computing

Abstract

This invention discloses a distributed collaborative control method and system for flexible distribution networks based on edge computing, belonging to the field of distribution network collaborative control technology. It collects operational data through edge computing nodes, calculates the upper limit of topology switching rate and energy regulation margin; enumerates topology combination schemes, including merging and decoupling schemes; performs multi-model simulation prediction using voltage fluctuation correlation models and energy mutual assistance control models; iteratively updates execution parameters, corrects voltage fluctuation and power regulation requirements, and derives dwell time; performs executability judgment based on voltage disturbance and power regulation margin constraints, generating a set of executable schemes; employs a hierarchical three-stage strategy to execute topology changes, reducing impact; monitors key parameters after execution, triggering a rollback mechanism to restore stability. The system includes a scheme generation module, a collaborative simulation module, a hierarchical execution module, and an autonomous recovery module. This invention reduces the impact of topology changes and improves the reliability and flexibility of the distribution network.

Description

Technical Field

[0001] This invention relates to the field of distribution network collaborative control technology, and more specifically, to a method and system for distributed collaborative control of flexible distribution networks based on edge computing. Background Technology

[0002] Existing flexible distribution network control methods largely rely on centralized master stations, which are computationally burdensome, have slow response times, and are ill-suited to the frequently fluctuating operating conditions in environments with high distributed power generation penetration. Centralized architectures pose a single point of failure risk during communication interruptions, and topology optimization often lacks precise assessment of the instantaneous impact of switching operations, easily leading to voltage exceeding limits or equipment overload. Furthermore, traditional methods lack effective online monitoring and rapid rollback mechanisms after control strategy execution, failing to ensure system safety when deviations occur between predictions and actual conditions, thus limiting the safety and flexibility of distribution network operation.

[0003] To address the above problems, this invention proposes a solution. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a distributed collaborative control method and system for flexible distribution networks based on edge computing, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A distributed collaborative control method for flexible distribution networks based on edge computing, including the following steps;

[0007] Step S1: Collect real-time operating data of each node in the flexible distribution network, calculate the upper limit of the current topology switching rate and the energy regulation margin, enumerate all possible merging and decoupling combination schemes based on the current network topology, form a candidate topology list, eliminate schemes that do not meet the switching frequency or regulation margin through constraint judgment, and output an acceptable candidate scheme set.

[0008] Step S2: Perform voltage fluctuation correlation model and energy mutual assistance control model simulation on the set of acceptable candidate schemes. The voltage model predicts the bus voltage step and the branch power flow surge, and the energy model deduces the power mutual assistance path between regions and the source-load-storage redistribution command. The prediction results are iteratively corrected by real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility judgment is completed, and the final executable scheme and its scheduling information are output.

[0009] Step S3: Formulate a graded execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, adopt a three-level progressive grid connection and adjust the power and topology step by step according to the ratio; otherwise, simplify the number of execution levels.

[0010] Step S4: Continuously monitor the key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compare the measured data with the predicted values, and immediately trigger the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

[0011] In a preferred embodiment, step S1 includes the following:

[0012] Calculate the current topology switching rate ceiling and energy regulation margin;

[0013] Topology switching rate limit This represents the maximum allowed topology switching frequency per unit time, determined based on the mechanical lifespan limitations of the switching equipment or the system's tolerance to frequent switching disturbances.

[0014] Energy regulation margin It represents the current redundancy capacity of the distribution network in terms of power and energy, and is calculated by statistically analyzing the remaining output margin and standby capacity of all adjustable resources;

[0015] Based on the current network topology, enumerate all possible combinations of merging and decoupling schemes, including:

[0016] Enumerate the merging and decoupling schemes. The merging scheme refers to connecting the originally electrically isolated network units together for operation, while the decoupling scheme refers to isolating the originally electrically connected parts for independent operation.

[0017] Schemes that do not meet the switching frequency or adjustment margin requirements are eliminated based on constraints, including: if a scheme's expected switching rate is greater than... This could lead to a power deficit exceeding [a certain threshold]. If it is, then mark it as unacceptable.

[0018] In a preferred embodiment, step S2 includes the following:

[0019] Simulations were performed on the acceptable candidate solution set using voltage fluctuation correlation model and energy mutual control model, including:

[0020] The voltage fluctuation correlation model is used to estimate the instantaneous voltage step amplitude and instantaneous power flow increment of each branch that may occur at each bus node after the candidate scheme is implemented.

[0021] The energy mutual assistance control model is used to deduce the output allocation and source-load-storage power redistribution commands of the flexible interconnected system under candidate schemes.

[0022] Iterative correction of prediction results using real-time data includes:

[0023] Voltage fluctuation amplitude correction uses the noise level of the current bus voltage and the short-time fluctuation background value to correct the voltage fluctuation prediction value;

[0024] Local power demand correction utilizes a real-time list of adjustable resources to correct active and reactive power adjustment commands, ensuring that they do not exceed the actual available adjustment capacity.

[0025] Dwell time derivation based on the topology switching rate limit Calculate the length of stay This indicates the minimum duration for which the system must maintain this topological state;

[0026] Feasibility assessment is performed based on the upper limit of voltage disturbance and the power regulation margin, including:

[0027] Voltage disturbance constraint: The corrected voltage fluctuation prediction value is compared with the upper limit of the voltage disturbance limit. If it is exceeded, it is determined that it is not feasible.

[0028] Power regulation margin constraint: This involves adjusting the active and reactive power regulation demands against the energy regulation margin. If the comparison exceeds the limit, the action is deemed unexecutable.

[0029] A scheme that simultaneously satisfies voltage disturbance constraints and power regulation margin constraints is marked as an executable scheme.

[0030] In a preferred embodiment, step S3 includes the following:

[0031] A tiered execution strategy is formulated based on the power balance difference and switching rate limitations, including:

[0032] Calculate the power balance difference , indicating the change in power flow before and after the connection point is connected to the grid;

[0033] like Exceeding the allowed single-step power step limit If so, a multi-stage gradual grid connection is adopted, and the power and topology are adjusted step by step according to the proportion; otherwise, the number of execution stages is simplified.

[0034] When using multi-stage gradual grid connection, each stage performs partial power regulation and topology changes, and there is at least a minimum waiting time between adjacent stages. = The system monitors the system response after each stage is completed, and triggers a protective shutdown if an anomaly is detected.

[0035] In a preferred embodiment, step S4 includes the following:

[0036] Continuous high-frequency monitoring of key bus voltage, tie-line power, controllable equipment output, and switching frequency; comparison of measured data with predicted values; and triggering conditions for the backoff mechanism include:

[0037] The measured voltage deviation of the critical busbar exceeded the allowable range.

[0038] Overload occurred on critical lines;

[0039] The control device outputs a value close to the limit threshold.

[0040] The edge computing-based distributed collaborative control system for flexible power distribution networks includes: a scheme generation module, a collaborative simulation module, a hierarchical execution module, and an autonomous recovery module, with signal connections between the modules.

[0041] Scheme generation module: Collects real-time operating data of each node in the flexible distribution network, calculates the upper limit of the current topology switching rate and energy regulation margin, enumerates all possible merging and decoupling combination schemes based on the current network topology, forms a candidate topology list, eliminates schemes that do not meet the switching frequency or regulation margin through constraint judgment, and outputs an acceptable candidate scheme set.

[0042] The collaborative simulation module performs voltage fluctuation correlation model and energy mutual assistance control model simulation on the acceptable candidate scheme set. The voltage model predicts the bus voltage step and branch power flow surge, and the energy model deduce the power mutual assistance path and source-load-storage redistribution command between regions. The prediction results are iteratively corrected through real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility is judged, and the final executable scheme and its scheduling information are output.

[0043] The tiered execution module formulates a tiered execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, a three-level progressive grid connection is adopted, and the power and topology are adjusted step by step in proportion; otherwise, the number of execution levels is simplified.

[0044] Autonomous Recovery Module: Continuously monitors key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compares the measured data with the predicted values, and immediately triggers the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

[0045] The technical effects and advantages of the edge computing-based distributed collaborative control method for flexible power distribution networks in this invention are as follows:

[0046] This invention introduces an edge computing architecture to distribute collaborative control tasks, significantly improving computational efficiency and response speed. By enumerating candidate topology schemes and combining parallel simulation prediction with voltage fluctuation and energy balance models, the feasibility and impact of switching operations can be accurately assessed before execution, effectively avoiding the execution of infeasible schemes. A hierarchical three-stage topology change execution strategy is adopted to ensure a smooth transition and greatly reduce the risk of transient impacts. At the same time, the system has real-time aftereffect monitoring and automatic rollback capabilities. Once an operational anomaly is detected, it can quickly recover to a stable state, thereby comprehensively improving the reliability, safety, and adaptability of flexible distribution network control. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the distributed collaborative control method for flexible power distribution networks based on edge computing according to the present invention.

[0048] Figure 2 This is a schematic diagram of the distributed collaborative control system module for flexible power distribution networks based on edge computing, as described in this invention. Detailed Implementation

[0049] 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, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0050] Please see Figure 1 As shown, this invention discloses a distributed collaborative control method for flexible distribution networks based on edge computing, including the following steps:

[0051] Step S1: Collect real-time operating data of each node in the flexible distribution network, calculate the upper limit of the current topology switching rate and the energy regulation margin, enumerate all possible merging and decoupling combination schemes based on the current network topology, form a candidate topology list, eliminate schemes that do not meet the switching frequency or regulation margin through constraint judgment, and output an acceptable candidate scheme set.

[0052] Step S2: Perform voltage fluctuation correlation model and energy mutual assistance control model simulation on the set of acceptable candidate schemes. The voltage model predicts the bus voltage step and the branch power flow surge, and the energy model deduces the power mutual assistance path between regions and the source-load-storage redistribution command. The prediction results are iteratively corrected by real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility judgment is completed, and the final executable scheme and its scheduling information are output.

[0053] Step S3: Formulate a graded execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, adopt a three-level progressive grid connection and adjust the power and topology step by step according to the ratio; otherwise, simplify the number of execution levels.

[0054] Step S4: Continuously monitor the key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compare the measured data with the predicted values, and immediately trigger the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

[0055] In step S1, real-time operating data of each node in the flexible distribution network is collected, the upper limit of the current topology switching rate and the energy regulation margin are calculated, all possible merging and decoupling combination schemes are enumerated based on the current network topology to form a candidate topology list, and schemes that do not meet the switching frequency or regulation margin are eliminated through constraint judgment, and an acceptable candidate scheme set is output. The specific contents include:

[0056] A flexible distribution network consists of several power supply areas, each of which includes a medium-voltage AC bus, distributed power generation devices, energy storage devices, and important loads within the area.

[0057] Each power supply area can physically operate as an independent power distribution unit, or it can achieve electrical mutual assistance through a flexible interconnection system. The flexible interconnection system consists of multi-port power electronic converters and controllable interconnection switches. The converters can perform bidirectional power distribution between different areas and allow flexible power coupling between AC and DC networks.

[0058] The open / closed state of each contact point determines whether the regions are decoupled and isolated for power supply or jointly connected to the grid for coordinated power supply. The contact points constitute the actual execution locations for subsequent topology merging and topology decoupling operations.

[0059] In this embodiment, the flexible distribution network maintains a predetermined power supply topology at the initial moment, in which each area independently undertakes local loads according to its own bus, key connection points are kept in predetermined open and closed states, energy storage is in a charged state that allows subsequent adjustment, and distributed power sources are kept within an adjustable output range that can be adjusted upwards or downwards. This initial topology and its corresponding voltage distribution, power flow direction, and power balance relationship serve as the baseline state for subsequent control steps.

[0060] By using sensors and edge computing nodes throughout the flexible power distribution network, the system collects real-time operational measurement data, including the voltage amplitude and phase of each bus node, the amplitude and direction of current flowing through each branch, the power injection and load power of each feeder, the output level and operating status of distributed power sources, the charging and discharging power and SOC level of energy storage devices, and the status of relevant circuit breaker switches. The data is then preprocessed and aggregated through the edge computing platform.

[0061] Two types of constraint parameters for calculating the current adjustability of a flexible distribution network include the upper limit of the topology switching rate. and energy regulation margin ;

[0062] This indicates the maximum permissible topology switching frequency per unit time, reflecting the mechanical lifespan limitations of the switching equipment or the system's tolerance to frequent switching disturbances. For example, if a mechanical switch has a limited number of safe operations per hour, then... It can be determined based on its cooling time and arc wear rate; for flexible electronic switches, The value is determined by the control response speed and the requirement to avoid voltage flicker. This value ensures that the subsequent scheme will not require too frequent switching operations. Take no more than once per minute, that is... The settings are based on the rated mechanical life of the on-site mechanical tie switch and the allowable frequency of voltage sag in the distribution network.

[0063] Adjustment margin This indicates the current redundancy capacity of the distribution network in terms of power and balance, including both power and energy levels. On the one hand, it involves statistically analyzing the remaining output margin and standby capacity of all adjustable resources to calculate the total active power margin that the system can increase and the total active load margin that can be reduced. Through the margin of each distributed power source The summation of accumulated loads and loads that can be flexibly unloaded is used to measure the ability to increase power supply capacity or reduce load capacity in a short period of time. On the other hand, considering the energy reserves of equipment such as energy storage batteries, the duration for which the load can be supported under islanded operation is assessed. These two aspects are combined to form... This is used to constrain whether the system's balancing capability is sufficient after a topology change. When the power adjustment required by a candidate solution exceeds the current... At that time, the plan will be considered infeasible;

[0064] Enumerate topology combinations: Obtain the current distribution network topology and equipment list, including the AC / DC backbone network connection relationships and the distribution locations of each flexible interconnection device;

[0065] Based on this topology model, we enumerate the possible merging and decoupling schemes that can be implemented under the current working conditions.

[0066] The combined scheme refers to connecting previously electrically isolated network units for operation, such as closing normally disconnected interconnection switches to form a network closed loop, or connecting two zones to the network through flexible interfaces such as MMC;

[0067] Decoupling and combination schemes refer to isolating the originally electrically connected parts and operating them independently. For example, disconnecting the connection point to make a subnet operate as an island, or cutting off the faulty area from the main network.

[0068] The enumeration of topology combinations is a combination search of the switch states of the distribution network, listing all possible sets of switch opening and closing state changes, but it must follow the basic operating constraints of the distribution network. In order to reduce the scheme space, it is preferable to consider only a finite number of topology changes, such as changing only one or two switch states in one operation, thereby avoiding overly complex combinations.

[0069] Decoupling combination schemes include disconnecting the TS switch to make a certain area operate completely in island mode, or disconnecting a certain area from the main grid and supplying it with backup power.

[0070] By exhaustively listing these combinations, a list of candidate network topology states is obtained. Each element in the list defines a set of switching actions to be performed and the resulting new topology.

[0071] For each candidate topology combination obtained by enumerating topology combinations, check whether the number and frequency of topology change operations involved in each combination meet the requirements. If a solution requires multiple consecutive switching within a very short period of time, and its total operation frequency exceeds the upper limit, then the solution will not be considered a feasible solution.

[0072] If any solution requires power support exceeding the rated capacity of the relevant flexible interface or exceeding the entire network... If the margin is insufficient, the plan is deemed unacceptable;

[0073] That is, if the expected switching rate of a certain merger plan is greater than Marked as unacceptable; if a decoupling scheme causes the feeder power and voltage gap to exceed If all conditions are met, mark it as unacceptable; if all conditions are met, mark it as acceptable.

[0074] Based on the above constraint judgment, combinations that do not meet the topology switching rate limit or have insufficient power regulation margin are removed from the candidate list. The remaining set of schemes ensures basic feasibility and provides a concise and high-quality candidate set for the next step of detailed simulation evaluation, outputting an acceptable candidate scheme set.

[0075] In step S2, voltage fluctuation correlation model and energy mutual assistance control model simulations are performed on the acceptable candidate scheme set. The voltage model predicts bus voltage step and branch power flow surge, and the energy model deduces inter-regional power mutual assistance paths and source-load-storage redistribution commands. The prediction results are iteratively corrected using real-time data. Based on the upper limit of voltage disturbance and power regulation margin, the feasibility is determined, and the final executable scheme and its scheduling information are output. Specific contents include:

[0076] For each candidate topology combination scheme in the acceptable candidate scheme set, multi-model simulation prediction is performed. The multi-model simulation prediction is completed in parallel by two types of models: voltage fluctuation correlation model and energy mutual control model.

[0077] The voltage fluctuation correlation model is used to estimate the instantaneous voltage step amplitude and instantaneous power surge of each branch node after the implementation of the candidate scheme; the energy mutual assistance control model is used to deduce the output distribution and source-load-storage power redistribution command of the flexible interconnection system under the candidate scheme, thereby deduce the local regulation demand.

[0078] The voltage fluctuation correlation model was established and deduced using a supervised learning approach. It was constructed as a set of pre-trained random forest regressors. Each random forest took a set of feature vectors as input and output two types of prediction indicators. The input feature vectors were obtained by concatenating the following information: First, the current operational measurement data, including the voltage amplitude and phase of each bus node, the current amplitude and direction of each branch, the injected power and load power of each feeder, the current active power output of distributed power sources and their adjustable or lowered regulation margins, and the current charging / discharging power and state of charge of energy storage devices; Second, the specific switching action sequence required by the candidate topology combination scheme, including the tie point numbers that need to be closed or opened, the order of actions, and the time interval between actions, thus providing the topology switching rate; Third, the operating status of each distributed power source at the expected switching time, including grid connection status, the allowed ramp rate of the inverter, and whether it is under power generation limitation constraints.

[0079] During training, the topology switching process observed in historical running events or simulation offline data is used as a sample. Each sample records the instantaneous step value of the bus voltage after the switch and the instantaneous surge value of the branch power flow at adjacent times, which are used as supervision labels respectively.

[0080] When performing prediction, the random forest outputs two quantities: the predicted value of the bus voltage step amplitude. And the predicted value of the sudden increase in branch power flow , The maximum difference between the voltage amplitude of each key bus and the steady-state voltage amplitude before the switch is completed in the first sampling period after the topology switch is completed, in volts or per unit. This represents the maximum increment of active power flow in each critical branch relative to the steady-state power flow before the topology switch, within the first sampling period after the switch, expressed in watts or kilowatts. The above definition ensures... and Both are quantifications of the degree of instantaneous disturbance, which can be directly used for subsequent constraint judgments;

[0081] Establishment and Derivation of the Energy Mutual Aid Control Model: Multiple feeders in a flexible distribution network are considered as multi-port objects coupled through a flexible interconnection system. The mutual aid paths and dynamic responses of power between different regions are characterized in state-space form. First, based on the network connection relationships under candidate topology combinations, the power transmission paths formed between regions via multi-port converters, tie points, and energy storage devices are extracted, resulting in the equivalent coupling matrix between regions. Second, each region is abstracted into three types of objects: source, load, storage, and power. The source corresponds to the adjustable output of distributed power sources, the load corresponds to the local load demand of the region, and the storage corresponds to the charging or discharging capacity of the energy storage devices in the region. The active power state and reactive power state of these three types of objects are jointly written into the state variable vector. The state update equation is given in linear discrete-time form. Output equation ;in, Indicates control input, The region output allocation command is represented by matrices A, B, and C, which are obtained by offline identification or small disturbance linearization. Their parameters are derived from historical operation records and field calibration data, including the power regulation slope of the converter under small signal conditions, the power ramp-up limit of energy storage under different states of charge, and the allowable power flow increment corresponding to the thermal stability limit of the feeder.

[0082] For each candidate topology combination, the state-space model is run within a short-time prediction window to obtain the region-level command prediction. and , After the candidate scheme is implemented, in order to keep the voltage and power flow of each bus within a safe range, the system needs to issue an active power adjustment command to each region. A positive value indicates that the active power injection of the region should be increased or the active power load of the region should be reduced, and a negative value indicates that the active power injection should be reduced or the transferable load absorption of the region should be increased. Defined as the reactive power adjustment command allocated to each region to maintain local voltage support after the candidate scheme is executed. A positive value indicates that capacitive reactive power support needs to be provided to the region, and a negative value indicates that reactive power needs to be absorbed.

[0083] By parallel extrapolating the two models described above, a set of feedforward prediction indices can be generated for each candidate topology combination scheme without actually switching physical switches. , , as well as This serves as input for subsequent iterative corrections and executability assessments;

[0084] Perform iterative parameter updates to correct the prediction results online and derive the residence time constraints of the scheme. The iterative parameter update consists of three parts;

[0085] The first part is the correction for voltage fluctuation amplitude, which is based on the voltage fluctuation correlation model. The current real-time measurement data is used for correction. The correction method is as follows: statistically analyze the noise level and short-time fluctuation background value of the current bus voltage. , It can be defined as the difference between the peak and valley values ​​of the bus voltage within the most recent sampling periods; Overlay Within the output confidence interval, the corrected voltage fluctuation prediction value is obtained. .like If it is significantly lower than the lower limit of field voltage noise, then... The model will be revised down to this lower limit to avoid overly optimistic predictions. This is the predicted voltage step amplitude of the scheme under the current operating conditions, which will be used for subsequent feasibility assessment.

[0086] The second part concerns local regulation power demand correction. This is based on the energy reconciliation control model. and The model is corrected using a real-time adjustable resource inventory. The correction method involves calculating the current adjustable output margins of distributed power sources, the upper limits of discharge and charge power for energy storage devices, and the unloading capacity of interruptible or transferable loads. These actual available values ​​are then compared with the model's predicted commands region by region. When the model's command exceeds the actual adjustable capacity of a region, the region is reclassified. and The maximum executable value in this region is truncated to obtain the corrected value. and This ensures that subsequent judgments use active and reactive power adjustment requirements that are still executable under on-site resource constraints.

[0087] The third part is the derivation of the length of stay. This is the minimum duration for which the system must maintain the topology state without further switching after executing the candidate topology combination scheme, in order to avoid high-frequency topology oscillations and protect the mechanical life of the switches. Based on the topology switching rate limit Calculations show that The aforementioned topology switching rate limit represents the maximum number of topology switching operations allowed per unit time, which has been determined in the preceding steps by the mechanical life constraints of the switching equipment, the control response characteristics of the flexible electronic interface, and the voltage flicker tolerance.

[0088] Define a lower bound for the dwell time for each candidate topology combination scheme. This is used to constrain subsequent scheduling: that is, once the solution is adopted and actually deployed to the site, its topology state must at least maintain [a certain state]. The duration of the switchover command must be specified, and no conflicting switchover command may be issued within that duration.

[0089] After the above three iterative updates, the corrected voltage fluctuation prediction value of the candidate scheme is obtained. Corrected local active power regulation demand Corrected local reactive power regulation demand and the residency time constraints tied to the scheme. ;

[0090] An executability assessment is performed to obtain the final set of executable solutions. The executability assessment is based on two types of constraints: voltage disturbance constraints and power regulation margin constraints.

[0091] Voltage disturbance constraints are used to determine whether the implementation of this scheme will cause an unacceptable step in bus voltage. An upper limit for voltage disturbance is set. , It is determined by the allowable instantaneous voltage deviation specifications of the distribution network, the tolerance of critical loads to voltage sags, and the voltage stabilization requirements of sensitive equipment on the user side;

[0092] Corrected voltage fluctuation prediction and If a comparison is made, Less than or equal to If the candidate solution satisfies the voltage perturbation constraint, then the candidate solution is considered to satisfy the voltage perturbation constraint; if Greater than If so, the candidate solution is deemed unfeasible due to its voltage impact;

[0093] Power regulation margin constraints are used to determine whether the system has sufficient regulation capability to maintain power flow balance after the scheme is implemented. Regulation margin It has been defined and calculated in the preceding steps. It also considers the adjustable generation margin, the adjustable generation margin, the unloading capacity of interruptible or transferable loads, and the dischargeable and absorbable energy of energy storage devices under the current state of charge. For each candidate scheme, the modified active power regulation requirements are... absolute value and Compare and simultaneously adjust the reactive power regulation requirements. absolute value and Compare. When absolute values Less than or equal to And absolute value Less than or equal to When the required active or reactive power regulation in any region exceeds the power regulation margin constraint, the candidate scheme is deemed to meet the power regulation margin constraint. If the candidate scheme is not feasible, it is marked as unworkable because it cannot achieve the required power flow redistribution and voltage support under the current available resources.

[0094] After considering both voltage disturbance constraints and power regulation margin constraints, a final binary decision result is obtained for each candidate scheme. If the candidate scheme simultaneously satisfies... Less than or equal to ,and and Not exceeding If the condition is met, the solution is marked as executable; otherwise, it is marked as non-executable.

[0095] For candidate schemes marked as executable, three types of scheduling information are further output: the first is the specific switching operation sequence of the executable feeder merging or decoupling scheme, which indicates the contact points that need to be closed or opened on site and the order of actions; the second is the minimum dwell time that the scheme must maintain after execution. The third is that the regional-level power regulation instruction set that energy storage devices and distributed power sources need to execute includes... and The allocation results in each region are used to directly distribute the data to the edge controller and field execution unit during actual deployment.

[0096] In step S3, a tiered execution strategy is formulated based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, a three-level progressive grid connection is adopted, adjusting the power and topology proportionally step by step; otherwise, the execution level is simplified, specifically including:

[0097] To avoid the impact of topology changes on the system, the control system designed a three-stage, hierarchical topology change execution strategy based on the parameter changes output in step S2. The magnitude and sequence of each stage of the change are determined by the previously calculated power balance difference. and switching rate limit Joint decision;

[0098] according to The size determines whether multi-level execution is needed. ;in, The current flow at the contact point before it was connected to the network. To predict the power flow at the same point after grid connection;

[0099] The maximum allowable single-step power step is: ;in, This is the rated transmission power of the connection line. This means that a single step voltage jump must not exceed 20% of the line's rated voltage to avoid thermal shock and transient voltage drops.

[0100] like If the power surge is large, a three-stage gradual grid connection will be adopted, with the first stage connecting to the total grid. 50% in the first phase, 30% in the second phase, and 20% in the third phase;

[0101] like If the power surge is small, it is simplified to one or two-stage operation.

[0102] according to Constraints on determining the timing intervals between operations at each level and the magnitude of each level: taking into account The given minimum interval = The controller will wait at least between two adjacent operation levels. After reaching a steady state, proceed to the next step.

[0103] The specific power and topology changes implemented at each stage can be divided according to a certain proportion; for example, the first stage implements approximately... The first stage executes 50%, the second stage executes the remaining 30%, and the third stage executes the last 20%. Alternatively, an unequal division can be adopted according to actual needs to better match the nonlinear response characteristics of the system.

[0104] Before each stage begins, the controller issues the target power or voltage command of the corresponding level to the device. The device adjusts its operation according to the command. When the stage is completed and the system indicators are stable, the next stage command is issued. Through such tiered control, the flexible interactive device acts as a buffer, enabling the change of topology to be completed in a soft transition manner, which greatly reduces voltage fluctuations and transient impacts.

[0105] Execute topology changes step by step and monitor responses: After developing a hierarchical execution plan, the control system begins physically operating switches to ensure that requirements are met. Allow time for topology changes; the specific process is as follows:

[0106] Implement Level 1 changes: The corresponding switches are put into or taken out as planned, and the relevant flexible equipment is controlled according to the Level 1 objectives. For example, if the plan is to connect the islanded area to the grid, the tie switch is put into operation at this time, but the converter is limited to a low level of initial grid connection power; if the plan is to transfer the load, the bypass switch is closed first and the supply side voltage is slightly increased.

[0107] After the first-level operation is completed, the monitoring of key parameter responses includes the voltage of each monitoring node. Power of connecting lines Actual changes, output power of important equipment And frequency changes, through high-speed data acquisition and local real-time analysis, determine whether the current system status remains within a safe range;

[0108] If everything is normal, then wait. Then continue to execute the second level change; if the monitoring detects an abnormality, the protective stop is immediately triggered. If necessary, the switch that was just put into operation can be disconnected and rolled back to the previous stable state. At this time, the rollback process in step S4 is entered.

[0109] The first stage was successfully completed. Then, the second and third stages of topology changes were carried out according to the pre-planned schedule: the switch and controller settings were adjusted to the new target values ​​one by one, and similar monitoring and verification were carried out at each step.

[0110] After the third stage is completed, all switches reach the target state, and the output of the flexible interface device also reaches the final required value. Then the topology change operation is completed, and the system enters a new steady-state operation mode.

[0111] In step S4, the voltage of the key bus, the power of the tie line, the output of controllable equipment, and the switching frequency are continuously monitored at high frequency. The measured data are compared with the predicted values. If the voltage deviation exceeds the limit, the line is overloaded, or the equipment output is close to the limit, the rollback mechanism is immediately triggered. After the rollback is completed, the process returns to step S1 for recalculation. The specific contents include:

[0112] After the topology change is completed and the system enters the new operating state, it continues to monitor key operating parameters at high frequency to promptly detect any discrepancies with simulation predictions. Key monitored parameters include: actual changes in voltage at each critical bus. Actual power of main connecting lines And the output power of adjustable devices involved in the control of the current direction. and no merit and topology switching frequency counting, among which, Used to calculate how much adjustment margin was actually utilized, in order to determine Is there any margin? The count is used to track whether the number of handovers that have occurred within a unit of time is close to the limit. Limitations. The monitoring process is collaboratively completed by edge control nodes distributed across various areas: local controllers acquire measured values ​​such as voltage and power in their respective areas and compare them with previously predicted values. , The data is compared with the predictions, and each node aggregates the monitoring data to the nearest edge computing center for global consistency checks. If the actual results deviate significantly from the predictions or violate the constraints, the next step of judgment is initiated.

[0113] Backoff mechanism triggering: Triggering conditions include measured critical bus voltage deviation exceeding the allowable range, overload of important lines, and control equipment output approaching the limit threshold;

[0114] Critical bus voltage deviation threshold: Trigger rollback; among which, 0.05 is the commonly used upper limit for short-time voltage steps on the distribution side, which can take into account both the tolerance of sensitive loads and the noise level on site, and the criterion is simple and clear;

[0115] Overload occurred on critical lines: ;in, When the power step deviation exceeds 10% of the line's rated value, it indicates that the operating condition is out of sync with the simulation, and continuing to execute it carries a high risk.

[0116] The control device output approaches the limit threshold: or Trigger rollback, reserve 5% dynamic margin to avoid control saturation leading to no gate for reverse adjustment or numerical integral drift;

[0117] If any triggering condition is met, the controller will first perform a rollback operation on the current topology according to the predetermined emergency strategy: for example, quickly disconnecting the newly closed interconnection switch to restore the network to the topology before the switchover, or cutting off part of the load as needed to ensure the stability of the main network;

[0118] The rollback operation can be completed in one step or in a multi-step sequence, depending on the specific circumstances. However, the principle is to prioritize restoring system stability rather than forcibly maintaining the new topology. In addition, the rollback process also emphasizes coordination with control strategy adjustments. If grid connection fails, the relevant converters are restored to islanded mode control to ensure that each region can operate independently and stably after the rollback. After the topology rollback is completed and the system stabilizes, the control flow automatically returns to step S1 to recalculate feasible candidate solutions based on the latest real-time data.

[0119] Please see Figure 2 As shown, this invention discloses a distributed collaborative control system for flexible power distribution networks based on edge computing, including: a scheme generation module, a collaborative simulation module, a hierarchical execution module, and an autonomous recovery module, with signal connections between the modules;

[0120] Scheme generation module: Collects real-time operating data of each node in the flexible distribution network, calculates the upper limit of the current topology switching rate and energy regulation margin, enumerates all possible merging and decoupling combination schemes based on the current network topology, forms a candidate topology list, eliminates schemes that do not meet the switching frequency or regulation margin through constraint judgment, and outputs an acceptable candidate scheme set.

[0121] The collaborative simulation module performs voltage fluctuation correlation model and energy mutual assistance control model simulation on the acceptable candidate scheme set. The voltage model predicts the bus voltage step and branch power flow surge, and the energy model deduce the power mutual assistance path and source-load-storage redistribution command between regions. The prediction results are iteratively corrected through real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility is judged, and the final executable scheme and its scheduling information are output.

[0122] The tiered execution module formulates a tiered execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, a three-level progressive grid connection is adopted, and the power and topology are adjusted step by step in proportion; otherwise, the number of execution levels is simplified.

[0123] Autonomous Recovery Module: Continuously monitors key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compares the measured data with the predicted values, and immediately triggers the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

[0124] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0125] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0126] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0127] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0128] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0129] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A flexible power distribution network distributed collaborative control method based on edge computing, characterized in that, Including the following steps: Step S1: Collect real-time operating data of each node in the flexible distribution network, calculate the upper limit of the current topology switching rate and the energy regulation margin, enumerate all possible merging and decoupling combination schemes based on the current network topology, form a candidate topology list, eliminate schemes that do not meet the switching frequency or regulation margin through constraint judgment, and output an acceptable candidate scheme set. Step S2: Perform voltage fluctuation correlation model and energy mutual assistance control model simulation on the set of acceptable candidate schemes. The voltage model predicts the bus voltage step and the branch power flow surge, and the energy model deduces the power mutual assistance path between regions and the source-load-storage redistribution command. The prediction results are iteratively corrected by real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility judgment is completed, and the final executable scheme and its scheduling information are output. Step S3: Formulate a graded execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, adopt a three-level progressive grid connection and adjust the power and topology step by step according to the ratio; otherwise, simplify the number of execution levels. Step S4: Continuously monitor the key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compare the measured data with the predicted values, and immediately trigger the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

2. The edge computing based flexible power distribution network distributed collaborative control method according to claim 1, characterized in that, Calculate the current topology switching rate ceiling and energy regulation margin; Topology switching rate limit This represents the maximum allowed topology switching frequency per unit time, determined based on the mechanical lifespan limitations of the switching equipment or the system's tolerance to frequent switching disturbances. Energy regulation margin It represents the redundancy capacity of the current distribution network in terms of power and energy, and is calculated by statistically analyzing the remaining output margin and standby capacity of all adjustable resources.

3. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 2, characterized in that, Based on the current network topology, enumerate all possible combinations of merging and decoupling schemes, including: Enumerate the merging and decoupling schemes. The merging scheme refers to connecting the originally electrically isolated network units together for operation, while the decoupling scheme refers to isolating the originally electrically connected parts for independent operation. Schemes that do not meet the switching frequency or adjustment margin requirements are eliminated based on constraints, including: if a scheme's expected switching rate is greater than... This could lead to a power deficit exceeding [a certain threshold]. If it is, then it is marked as unacceptable.

4. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 1, characterized in that, Simulations were performed on the acceptable candidate solution set using voltage fluctuation correlation model and energy mutual control model, including: The voltage fluctuation correlation model is used to estimate the instantaneous voltage step amplitude and instantaneous power flow increment of each branch that may occur at each bus node after the candidate scheme is implemented. The energy mutual assistance control model is used to deduce the output allocation and source-load-storage power redistribution commands of the flexible interconnected system under candidate schemes.

5. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 4, characterized in that, Iterative correction of prediction results using real-time data includes: Voltage fluctuation amplitude correction uses the noise level of the current bus voltage and the short-time fluctuation background value to correct the voltage fluctuation prediction value; Local power demand correction utilizes a real-time list of adjustable resources to correct active and reactive power adjustment commands, ensuring that they do not exceed the actual available adjustment capacity. Dwell time derivation based on the topology switching rate limit Calculate the length of stay This represents the minimum duration for which the system must maintain this topological state.

6. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 4, characterized in that, Feasibility assessment is performed based on the upper limit of voltage disturbance and the power regulation margin, including: Voltage disturbance constraint: The corrected voltage fluctuation prediction value is compared with the upper limit of the voltage disturbance limit. If it is exceeded, it is determined that it is not feasible. Power regulation margin constraint: This involves adjusting the active and reactive power regulation demands against the energy regulation margin. If the comparison exceeds the limit, the action is deemed unexecutable. A scheme that simultaneously satisfies voltage disturbance constraints and power regulation margin constraints is marked as an executable scheme.

7. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 1, characterized in that, A tiered execution strategy is formulated based on the power balance difference and switching rate limitations, including: Calculate the power balance difference , indicating the change in power flow before and after the connection point is connected to the grid; like Exceeding the allowed single-step power step limit If the grid connection is not ideal, a multi-stage gradual grid connection is adopted, and the power and topology are adjusted step by step according to the proportion; otherwise, the number of execution stages is simplified.

8. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 7, characterized in that, When using multi-stage gradual grid connection, each stage performs partial power regulation and topology changes, and there is at least a minimum waiting time between adjacent stages. = The system monitors the system response after each stage is completed, and triggers a protective shutdown if an anomaly is detected.

9. The distributed collaborative control method for flexible distribution networks based on edge computing according to claim 1, characterized in that, Continuous high-frequency monitoring of key bus voltage, tie-line power, controllable equipment output, and switching frequency; comparison of measured data with predicted values; and triggering conditions for the backoff mechanism include: The measured voltage deviation of the critical busbar exceeded the allowable range. Overload occurred on critical lines; The control device outputs a value close to the limit threshold.

10. A distributed collaborative control system for flexible distribution networks based on edge computing, used to implement the distributed collaborative control method for flexible distribution networks based on edge computing as described in any one of claims 1-9, characterized in that, include: Scheme generation module: Collects real-time operating data of each node in the flexible distribution network, calculates the upper limit of the current topology switching rate and energy regulation margin, enumerates all possible merging and decoupling combination schemes based on the current network topology, forms a candidate topology list, eliminates schemes that do not meet the switching frequency or regulation margin through constraint judgment, and outputs an acceptable candidate scheme set. The collaborative simulation module performs voltage fluctuation correlation model and energy mutual assistance control model simulation on the acceptable candidate scheme set. The voltage model predicts the bus voltage step and branch power flow surge, and the energy model deduce the power mutual assistance path and source-load-storage redistribution command between regions. The prediction results are iteratively corrected through real-time data. Based on the voltage disturbance upper limit and power regulation margin, the feasibility is judged, and the final executable scheme and its scheduling information are output. The tiered execution module formulates a tiered execution strategy based on the power balance difference and switching rate limit. If the power surge exceeds the set limit, a three-level progressive grid connection is adopted, and the power and topology are adjusted step by step in proportion; otherwise, the number of execution levels is simplified. Autonomous Recovery Module: Continuously monitors key bus voltage, tie line power, controllable equipment output and switching frequency at high frequency, compares the measured data with the predicted values, and immediately triggers the rollback mechanism if the voltage deviation exceeds the limit, the line is overloaded or the equipment output is close to the limit. After the rollback is completed, the process returns to step S1 to recalculate.

Images