A cloud edge collaboration-based backup power automatic input control method and system

Abstract

The application discloses a kind of based on cloud edge cooperation's backup power automatic input control method, it is applied to including station end execution device and cloud edge cooperation main station's cloud edge cooperation backup automatic switching system, cloud edge cooperation main station obtains whole network real-time data update power grid topology model, obtains each station end current backup strategy and operating data in power grid topology model Simulation backup action, the security and stability of power flow distribution is checked, and it is recalculated optimal strategy if not pass, and it is used current strategy if pass, generate dynamic strategy table and send to corresponding station end execution device;Station end execution device real-time monitoring and the communication link state of cloud edge cooperation main station, communication is normal when receiving and buffering dynamic strategy table, when communication interruption, load last buffered dynamic strategy table or local strategy table, through FPGA execution fault detection and jump brake operation.The application realizes millisecond level response, guarantees strategy global optimal again, and independently reliable operation when communication is abnormal.

Description

Technical Field

[0001] This invention relates to power system automation technology, specifically to a method and system for automatic switching control of backup power supply based on cloud-edge collaboration. Background Technology

[0002] Automatic transfer switching (ATS) is a key automatic device in power systems to ensure power supply reliability. When the main power supply is disconnected due to a fault, ATS can quickly put the backup power supply into operation to shorten the power outage time and maintain power supply to important loads. Currently, there are two main types of ATS implementation methods, but both present inherent contradictions that are difficult to reconcile.

[0003] The first type is the substation-based local automatic transfer switch (ATS). These devices are deployed directly within the substation, making judgments based on local electrical quantities and implementing tripping and closing via hard contacts or dedicated logic circuits. Their action speed can reach millisecond levels, meeting the stringent time requirements for rapid fault clearing and power restoration. However, because the device only collects local information, its operational logic is fixed at the factory and cannot adaptively correct itself based on changes in grid topology or operational modes during operation. This "information silo" working mode prevents the device from perceiving the overall grid status, frequently leading to the following problems in actual operation: when the grid operation mode is adjusted (e.g., line maintenance, load transfer), the fixed strategy may no longer be applicable; in multi-source power supply or ring network structures, the local ATS action may cause overload of the backup power supply in adjacent substations, or even cause uncoordinated actions with other safety automatic devices such as stabilization devices and low-frequency load shedding. Therefore, although the substation-based ATS is "fast," it lacks intelligent perception and coordination capabilities regarding the overall operational status, often summarized in the industry as "fast but not intelligent."

[0004] The second type is the centralized master station backup automatic transfer system. This type of solution deploys the decision-making logic at the dispatch center master station, using network-wide information collected by the SCADA / EMS system for unified modeling and strategy optimization, theoretically achieving global optimization of the backup transfer strategy. However, the centralized master station solution has two inherent drawbacks: First, the action chain is long. From the occurrence of a fault, information transmission, master station decision-making, command issuance to execution at the station end, the entire process is affected by factors such as communication latency, master station computing load, and human-machine confirmation, resulting in action speeds typically on the order of seconds or even minutes, which cannot meet the stringent millisecond-level action requirements of backup automatic transfer. Second, survivability is poor. Once the communication link between the station end and the master station is interrupted due to natural disasters, equipment failures, or network attacks, the station end will completely lose its backup automatic transfer function, falling into a paralyzed state of "no strategy available." Therefore, although the centralized master station solution has "superior strategy," it sacrifices speed and independent survivability, and is often summarized in the industry as "intelligent but not fast."

[0005] In recent years, cloud-edge collaborative technology has been gradually applied in fields such as power system dispatching and operation, and centralized monitoring of new energy sources, such as the cloud-edge integrated intelligent dispatching and operation platform built by China Southern Power Grid. However, objectively speaking, these existing applications are mainly geared towards dispatching optimization, status monitoring, and data analysis scenarios at the minute or even hour level. Their system architecture, communication protocol stacks (such as general IoT protocols), and control logic design have not been optimized for millisecond-level fast protection control. More importantly, existing cloud-edge collaborative solutions generally assume continuous availability of communication links and rarely consider functional survivability issues under extreme conditions such as communication interruptions. Their "failure upon network outage" architectural characteristic fundamentally conflicts with the rigid requirement of "reliable operation during faults" for automatic transfer switches.

[0006] Based on the above analysis, how to integrate the speed of station-side control with the global nature of master station control in the specific field of automatic transfer switch (ATS) equipment, and construct a novel architecture that can achieve millisecond-level fault response, ensure global policy optimization, and operate independently and reliably during communication anomalies, has become a pressing technical challenge in this field. Currently, no mature solution has emerged that can simultaneously meet all three requirements. Summary of the Invention

[0007] The technical problem to be solved by this invention: (1) Although the local automatic transfer switch at the station can operate in milliseconds, the strategy is fixed and cannot perceive the overall power grid status, which may easily cause the backup power supply to overload or be incompatible with other safety automatic devices. (2) Although the centralized backup automatic transfer strategy of the main station is the best overall strategy, the speed of action is constrained by communication delay, and the function is paralyzed when communication is interrupted. (3) Existing cloud-edge collaborative technologies are geared towards long-cycle scheduling, which cannot meet the requirements of backup self-starting in milliseconds, nor do they consider functional survivability under extreme conditions such as communication interruption.

[0008] To address the aforementioned problems in existing technologies, this paper provides a method and system for automatic backup power supply control based on cloud-edge collaboration. This system aims to resolve the technical contradiction between the "fast but not intelligent" nature of local backup power supply at the station and the "intelligent but not fast" nature of centralized backup power supply at the master station. It achieves millisecond-level response, ensures global optimal strategy, and operates independently and reliably during communication anomalies, filling the application gap of cloud-edge collaboration technology in the field of rapid protection and control.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: An automatic backup power supply control method based on cloud-edge collaboration is applied to a cloud-edge collaborative automatic backup power supply system, which includes a station-side execution device and a cloud-edge collaborative master station. The station-side execution device is deployed in the substation, and the cloud-edge collaborative master station is deployed in the dispatch center. The method includes the following steps: The cloud-edge collaborative master station acquires real-time data from the entire network according to a preset verification cycle or when a change in operating mode is detected. It updates the power grid topology model using the real-time data from the entire network, acquires the current backup strategy and operating data of each station-end execution device, simulates backup actions in the power grid topology model, calculates the power flow distribution of the corresponding power station after backup, and performs a safety and stability verification based on the power flow distribution. For station-end execution devices that fail the verification, the optimal backup strategy is recalculated. For station-end execution devices that pass the verification, the current backup strategy is retained. A dynamic strategy table containing fault criterion parameters and action sequences corresponding to the backup strategy is generated and sent to the corresponding station-end execution device. The station-side execution device monitors the communication link status with the cloud-edge collaborative master station in real time. When the communication is normal, it receives and caches the dynamic policy table. When the communication is interrupted, it loads the last cached dynamic policy table or the preset local policy table. The FPGA hardware module in the station-side execution device performs fault detection based on the fault criterion parameters in the dynamic policy table or the local policy table and the real-time collected electrical quantity data of the station. When a fault is detected, it independently executes tripping and closing operations according to the action sequence in the loaded dynamic policy table or the local policy table.

[0010] Furthermore, when performing safety and stability verification based on the power flow distribution, the step of performing cascading overload prevention verification is included. Specifically, after simulating backup operation, the backup power load rate of all adjacent substations that have a power supply relationship with the substation where the station-end execution device is located is verified. If the backup power load rate of any adjacent substation exceeds the preset ratio of its maximum allowable load rate, it is determined that there is a risk of cascading overload, and the verification fails.

[0011] Furthermore, when performing safety and stability verification based on the power flow distribution, the process includes a step of multi-station collaborative conflict detection verification. Specifically, for substations where multiple station-end execution devices are located in a chain or ring network power supply structure, a directed graph of the action sequence of each substation is constructed, and a depth-first search algorithm is used to detect whether there are conflict nodes where multiple substations simultaneously compete for the same backup power supply. If such nodes exist, the verification fails.

[0012] Furthermore, when performing safety and stability verification based on the power flow distribution, the step of performing transient stability pre-verification is included. Specifically, the transient process within a preset time after the simulated backup action is pre-evaluated, and the system frequency deviation and critical node voltage transient drop after backup are verified to see if they exceed preset thresholds. If they do, the verification fails.

[0013] Furthermore, when recalculating the optimal backup strategy for substation execution devices that fail the verification, specifically for these devices, an objective function is established with multiple objectives, including minimizing weighted load shedding losses, backup power reliability penalties, number of switching actions, and voltage quality deviations. Backup power capacity constraints, voltage constraints, topology connectivity constraints, number of switching actions constraints, cascaded safety constraints, and transient safety constraints are also established. The optimal backup strategy is then obtained by solving these constraints. Specifically, the cascaded safety constraint ensures that the backup power load rate of all adjacent substations after backup does not exceed a preset proportion of their maximum allowable load rate. The transient safety constraint ensures that the system frequency deviation within a preset time after backup action does not exceed a preset frequency threshold, and the transient voltage drop at critical nodes does not fall below a preset voltage threshold.

[0014] Furthermore, after the station-side execution device monitors the communication link status with the cloud-edge collaborative master station in real time, it also includes: when communication is restored, the station-side execution device reports the event records during the communication interruption to the cloud-edge collaborative master station. The cloud-edge collaborative master station uses real-time data from the entire network to update the power grid topology model, obtains the current backup strategy and operating data of the station-side execution device, simulates backup actions in the power grid topology model, calculates the power flow distribution of the corresponding power station after backup, and performs a safety and stability check based on the power flow distribution. If the check fails, the optimal backup strategy is recalculated; if the check passes, the current backup strategy is retained, a dynamic strategy table is generated, and sent to the station-side execution device. The first strategy message sent by the cloud-edge collaborative master station to the station-side execution device after communication is restored carries a strategy change flag during the interruption.

[0015] Furthermore, when the FPGA hardware module in the station-side execution device performs fault detection based on the fault criterion parameters in the dynamic strategy table or the local strategy table and the real-time collected local electrical quantity data, specifically, the FPGA hardware module substitutes the fault criterion parameters and the local electrical quantity data into the fault criterion formula for calculation. The fault criterion formula is as follows:

[0016] in, This refers to the measured effective value of the bus line voltage in the electrical quantity data of this station; The voltage criterion threshold is one of the fault criterion parameters; This refers to the measured effective value of the incoming phase current in the electrical quantity data of this station; The current criterion threshold is one of the fault criterion parameters; The continuous duration during which the bus voltage is lower than the voltage criterion threshold and the incoming line current is lower than the current criterion threshold; The time setting is used as a parameter in the fault diagnosis criteria. This indicates that a fault has been detected. This indicates that no fault was detected.

[0017] Furthermore, if the fault criterion parameters are those in the dynamic strategy table, the voltage criterion threshold is the result of multiplying the voltage value of the target bus under the minimum allowable operating conditions after backup power is applied by a preset safety factor, and the current criterion threshold is the result of multiplying the current value of the working power supply incoming line under the minimum load conditions by a preset reliability factor. The time setting is determined comprehensively based on the switching action sequence requirements and the coordination time of the line protection. If the fault criterion parameters are those in the local strategy table, the voltage criterion threshold is the product of the rated voltage and the corresponding coefficient, the current criterion threshold is the product of the rated current and the corresponding coefficient, and the time setting is adjusted within a specified time range.

[0018] Furthermore, the changes in operation mode include: changes in the open / closed state of circuit breakers or disconnectors associated with the substation where the station-end execution device is deployed; switching of the wiring method of the busbar where the substation is located; or the commissioning or decommissioning of tie lines or feeders related to the substation.

[0019] Furthermore, each time the station-side execution device performs fault detection and trip / close operations, it records the policy identifier of the currently used dynamic policy table or local policy table to the operation log and reports it to the cloud-edge collaborative master station along with the event report. The cloud-edge collaborative master station determines the policy version currently used by the station-side execution device by comparing the policy identifier reported by the station-side execution device with the policy identifier of the latest issued dynamic policy table.

[0020] The present invention also proposes a cloud-edge collaborative automatic transfer system, comprising a station-end execution device and a cloud-edge collaborative master station. The station-end execution device is deployed within the substation, and the cloud-edge collaborative master station is deployed in the dispatch center. The cloud-edge collaborative automatic transfer system is programmed or configured to implement any of the cloud-edge collaborative automatic transfer control methods for backup power supply.

[0021] Compared with the prior art, the advantages of the present invention are as follows: (1) To address the issue of fixed local backup automatic transfer strategy at the station end, this invention obtains real-time data from the entire network through cloud-edge collaboration with the master station to update the power grid topology model, performs security and stability checks, recalculates the optimal backup transfer strategy, and generates a dynamic strategy table containing fault criterion parameters and action sequences, which is then sent to the station end. The station end FPGA hardware module performs fault detection and trip / close operations based on this dynamic strategy table. Thus, while retaining millisecond-level fast action at the station end, the fault criterion parameters and action sequences are dynamically optimized by the master station based on the global state, avoiding strategy fixation and incoordination between devices caused by information silos.

[0022] (2) To address the issue of slow speed and reliance on communication in centralized automatic transfer switching at the main station, this invention centralizes safety and stability verification and strategy optimization at the main station side, using a preset verification cycle or detection of changes in operating mode as trigger conditions for strategy optimization; while fault detection and tripping / closing operations are independently executed by the station-side FPGA hardware module according to the loaded dynamic strategy table or local strategy table, without relying on real-time communication at the main station. This decouples global strategy optimization from millisecond-level fault handling, and communication delay no longer affects the speed of fault handling.

[0023] (3) To address the issue that existing cloud-edge collaborative technologies do not consider survivability during communication interruptions, this invention monitors the communication link status in real time at the station. When communication is normal, it receives and caches the dynamic policy table issued by the master station. When communication is interrupted, it automatically loads the last cached dynamic policy table or the preset local policy table, and the FPGA hardware module continues to independently perform fault detection and trip / close operations. This achieves both policy optimization under normal operating conditions and functional availability under communication anomalies. Attached Figure Description

[0024] Figure 1 This is a flowchart of a method according to an embodiment of the present invention.

[0025] Figure 2 This is a detailed flowchart of the cloud-edge collaborative master station collaborative verification process according to an embodiment of the present invention.

[0026] Figure 3 This is a detailed flowchart of the operation of the station-side execution device according to an embodiment of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and specific preferred embodiments, but this does not limit the scope of protection of the present invention.

[0028] Before introducing specific embodiments of the present invention, relevant concepts or terms will be explained.

[0029] Pre-set strategy: refers to the default action logic burned into the device at the factory, which is used as a local strategy table.

[0030] Dynamic policy table: refers to the policies generated and distributed by the cloud-edge collaboration master station and stored in the station-side policy cache module. It is the main basis for actions in the collaboration mode.

[0031] Last effective strategy: Specifically refers to the latest dynamic strategy table cached on the station side before communication is interrupted.

[0032] Example This embodiment proposes a cloud-edge collaborative automatic backup power supply control method. This method is applied to a cloud-edge collaborative automatic backup power supply system, which includes a station-side execution device and a cloud-edge collaborative master station. The station-side execution device and the cloud-edge collaborative master station are connected via a power dispatch data network. Wherein: The station-side execution device is deployed within a 110kV or 10kV substation and employs an embedded hardware platform. The station-side execution device collects local electrical quantity data in real time and monitors the communication link status L with the master station, automatically switching operating modes based on the link status L. Entering collaborative mode at ) When the fault condition is met, the system enters the independent operation mode. In the collaborative mode, it obtains the dynamic policy table issued by the cloud-edge collaborative master station. In the independent operation mode, it obtains the preset fixed policy or the last effective policy. When the fault condition is met, it independently executes the trip and close operation within a few hundred milliseconds.

[0033] The core hardware of the station-side execution device includes: Main control CPU module: Adopts industrial-grade ARM Cortex-A series processor, responsible for running software functions such as communication protocol stack, dynamic policy table parsing, data cache management, communication status monitoring and mode switching.

[0034] Data acquisition module: It consists of analog input circuit, low-pass filter, high-precision ADC (sampling rate not less than 4kHz) and opto-isolated input circuit, realizing the synchronous acquisition of bus voltage RMS value U, incoming current RMS value I and switch status S.

[0035] FPGA coprocessor module: Used for hardware acceleration of millisecond-level fast logic, with internally embedded fault detection algorithms, directly outputting trip and close pulses to ensure that the action time does not exceed 100ms. This module reads fault criterion parameters and action sequences from a designated partition of the eMMC flash memory, independently completing fault detection and trip / close execution, without relying on CPU intervention throughout the entire process.

[0036] Storage modules include DDR3 memory and eMMC flash memory. The eMMC flash memory is divided into the following partitions: firmware area (stores pre-configured policies), dynamic policy cache area (stores dynamic policy tables issued by the master station), and log area (stores operation events and fault waveforms).

[0037] Communication module: Supports dual Ethernet for communication with the main station; also has RS485 / fiber optic interface for intra-station interaction.

[0038] Power module: Supports AC and DC dual power supply and has power-off retention function.

[0039] The cloud-edge collaborative master station is deployed in the dispatch center, based on general-purpose server hardware, and adopts a distributed architecture. The cloud-edge collaborative master station constructs a power grid topology model and periodically verifies the security and stability of the backup strategies currently used by each station. When the verification fails or the operating mode changes, the optimal backup scheme is recalculated, a dynamic strategy table is generated, and the generated dynamic strategy table is sent to the corresponding station.

[0040] The cloud-edge collaborative main station mainly includes: Data server: runs a real-time database with a data refresh cycle of no more than 2 seconds.

[0041] Application server: Deploys core algorithm modules such as online verification and strategy optimization to meet the requirements of second-level verification and strategy calculation.

[0042] Historical server: Used to store historical data.

[0043] Front-end communication server: responsible for establishing communication connections with each station, managing heartbeats and policy distribution.

[0044] like Figure 1 As shown, the method in this embodiment includes the following steps: S101: Cloud-edge collaborative main station collaborative verification process.

[0045] The cloud-edge collaborative master station acquires real-time data from the entire network according to a preset verification cycle or when a change in operating mode is detected. It updates the power grid topology model using the real-time data from the entire network, acquires the current backup strategy and operating data of each station-end execution device, simulates backup actions in the power grid topology model, calculates the power flow distribution of the corresponding power station after backup, and performs a safety and stability verification based on the power flow distribution. For station-end execution devices that fail the verification, the optimal backup strategy is recalculated. For station-end execution devices that pass the verification, the current backup strategy is retained. A dynamic strategy table containing fault criterion parameters and action sequences corresponding to the backup strategy is generated and sent to the corresponding station-end execution device. S102: Station-side execution device operation process.

[0046] The station-side execution device monitors the communication link status with the cloud-edge collaborative master station in real time. When the communication is normal, it receives and caches the dynamic policy table. When the communication is interrupted, it loads the last cached dynamic policy table or the preset local policy table. The FPGA hardware module in the station-side execution device performs fault detection based on the fault criterion parameters in the dynamic policy table or the local policy table and the real-time collected electrical quantity data of the station. When a fault is detected, it independently executes tripping and closing operations according to the action sequence in the loaded dynamic policy table or the local policy table.

[0047] Through the above steps, the following three mechanisms work together: Mechanism 1 (Millisecond-level rapid tripping and closing): Regardless of the mode, the FPGA hardware module at the station always uses local data and built-in criteria to achieve independent and rapid fault handling within hundreds of milliseconds.

[0048] Mechanism 2 (Second-level online verification and dynamic policy update): During normal communication, the master station performs a global verification at a second-level cycle, dynamically optimizes the policy, and sends it to the station-side cache to achieve "dynamic optimization of the policy".

[0049] Mechanism 3 (Independent Operation During Communication Interruption): When communication is interrupted, the station automatically switches to independent operation mode, loads the last valid policy in the cache or the preset fixed policy to ensure that basic functions are not invalidated; after communication is restored, it automatically switches back to collaborative mode.

[0050] The above three mechanisms form a complete closed-loop control logic of "normal optimization, rapid fault response, and self-recovery in case of anomalies". The collaborative workflow is summarized as follows: Routine operation (mechanism 2 as the main force): The main site periodically checks and dynamically optimizes the strategy, and sends it to the site to update the cache.

[0051] Fault occurrence (mechanism 1 dominant): The station monitors the fault and immediately performs tripping and closing, without relying on the master station.

[0052] Communication Anomaly (Mechanism 3 Dominant): Upon detecting a communication interruption, the station immediately switches to independent mode, using the last effective strategy or a pre-defined strategy. If a fault occurs at this time, Mechanism 1 will still be executed. After communication is restored, it automatically switches back to collaborative mode and reports the event.

[0053] The following provides a detailed explanation of each step.

[0054] like Figure 1 As shown, the cloud-edge collaborative backup self-connection master station collaborative verification process in step S101 starts with two parallel triggering conditions. One is the periodic verification periodic trigger, that is, the master station performs verification according to the preset verification period. (Adjustable from 5 to 60 seconds) The verification process is initiated periodically; the second is the detection of changes in operating mode, that is, the master station continuously monitors the switch status changes, bus operation mode changes, and line commissioning / decommissioning events related to each standby automatic transfer station (i.e., substations with deployed station-side execution devices) in the power grid topology model. When a change in the topology corresponding to any station is detected, the station is immediately marked as "change in operating mode" and the standby strategy for that station is re-verified first within the current verification cycle, without waiting for the next regular verification cycle.

[0055] In this embodiment, changes in operating mode include: changes in the open / closed state of circuit breakers or disconnectors associated with the substation where the station-end execution device is deployed; switching of the wiring method of the busbar where the substation is located; or the commissioning or decommissioning of tie lines or feeders related to the substation. The specific judgment logic is as follows: (a) Change in switch status - When the open / closed status of a circuit breaker or disconnector associated with a backup automatic transfer station in the topology model changes, the change in operating mode is determined; (b) Change of busbar operation mode - When the wiring method of the busbar where the station is located (such as single busbar or double busbar segmentation) is switched, the change of operation mode is determined; (c) Line commissioning / decommissioning events – When a site-related tie line or feeder is commissioned or decommissioned, the change in operating mode is determined.

[0056] Regardless of the triggering condition, the cloud-edge collaborative backup self-distribution master station enters the subsequent processing flow. For example... Figure 1 As shown, the subsequent processing flow includes: S201) Data Preparation: The cloud-edge collaborative backup self-connection master station obtains real-time data of the entire network from the EMS system, including the power flow section and station-end parameters of the entire network, and uses this data to update the power grid topology model. The power grid topology model is built based on the CIM standard.

[0057] S202) Backup power simulation: The cloud-edge collaborative backup power master station obtains the backup power strategy and operation data currently adopted by the execution device at each station. In the updated power grid topology model, it simulates the backup power action of each station at each backup power station. Specifically, it simulates the power supply loss after the working power supply loses voltage, puts the backup power supply into operation according to the current backup power strategy, calculates the power flow after the switch, and obtains the power flow distribution of the power station after the backup power is put into operation.

[0058] S203) Constraint Verification: The cloud-edge collaborative backup self-connection master station performs a security and stability check based on the calculated power flow distribution. If all checks pass, the corresponding station-side execution device maintains the existing strategy; otherwise, strategy optimization is triggered. This verification includes six dimensions: (1) Overload check: Calculate the load rate of the standby power supply after it is put into operation. Determine whether it exceeds the maximum allowable load rate. .like ( If the value is between 0.9 and 1.0, the verification will fail. The backup power load rate after the backup power supply is put into operation. The current load of the backup power supply. For the load to be switched, Rated capacity of backup power supply This represents the maximum allowable load rate.

[0059] (2) Voltage verification: Determine the voltage of all nodes after backup power is applied. Is it in Inside( , If any limit is exceeded, the test will fail. For post-investment milestones voltage amplitude, and These are the lower and upper limits of the permissible voltage, respectively.

[0060] (3) Topology verification: The depth-first search (DFS) algorithm is used to check whether there are unplanned islands or electromagnetic loops in the power grid after the backup is put into operation. If they exist, the test fails.

[0061] (4) Cascading Overload Prevention Verification: This verification is not limited to checking whether the standby power supply of this substation is overloaded, but also extends to checking the standby power supply load rate of all adjacent substations that have a power supply relationship with this substation. After simulating the standby switching action, the standby power supply load rate of all adjacent substations that have a power supply relationship with the substation where the station-end execution device is located is further verified. If the standby switching action causes the standby power supply load rate of any adjacent substation to exceed its maximum allowable load rate preset ratio (e.g., 80%), thereby exceeding the warning threshold, then it is determined that there is a risk of cascading overload, and the verification fails. This mechanism can effectively prevent a single standby automatic transfer action from causing a regional chain overload.

[0062] (5) Multi-station collaborative conflict detection and verification: For multiple automatic transfer stations (i.e., substations where the station-side execution devices are located) in a chain or ring network power supply structure, the system verifies whether there are any potential timing conflicts in the backup power transfer actions of multiple stations. Specifically, a directed graph of the action sequence of each automatic transfer station is constructed, and the DFS algorithm is used to detect whether there are conflicting nodes where multiple substations are simultaneously competing for the same backup power supply. If such a conflict exists, the verification fails, and the conflict information is fed back to the cloud-edge collaborative master station for coordination and optimization. This mechanism ensures that the backup power transfer strategies of multiple stations under the cloud-edge collaborative architecture are coordinated and consistent at the global level.

[0063] (6) Transient stability pre-verification after automatic backup switching: Based on the power grid topology model and typical fault scenarios, a rapid pre-assessment is performed on the transient process within a preset time (e.g., 0~500ms) after the simulated automatic backup switching action, verifying whether the system frequency deviation and voltage transient drop at key nodes are within the allowable range. A simplified transient stability criterion is adopted (frequency deviation Δf ≤ 0.5Hz and voltage drop amplitude ≤ 15% of the rated value, duration ≤ 200ms). If any condition is exceeded, the verification fails and strategy optimization is triggered.

[0064] S204) Strategy Optimization: After the above six-dimensional verification, the cloud-edge collaborative master station determines whether all verifications have passed. If all verifications pass, the current backup strategy is used; if any verification fails, the optimal backup strategy is recalculated.

[0065] When recalculating the optimal backup power supply strategy, a multi-objective optimization method is adopted. Specifically, the following objective function is established with the multiple objectives of minimizing weighted load shedding losses, backup power supply reliability penalties, number of switching operations, and voltage quality deviation: min F = α1· Σ(w i · ΔP i ) + α2· R(S_backup) + α3· N_sw + α4· V_penalty Where: the first term Σ(w) i · ΔP i ) represents the weighted load shedding loss, w i Assign importance weights to load i (Level 1 load w=10, Level 2 load w=3, Level 3 load w=1), ΔP i The first term is the load shedding amount (MW) of load i; the second term, R(S_backup), is the backup power supply reliability penalty, R(S_backup) = R0·exp(ρ / S_rated), where ρ is the current load rate of the backup power supply, S_rated is the rated capacity of the backup power supply, and R0 is the basic penalty coefficient; the third term, N_sw, is the total number of switching operations, introduced to extend the life of switching equipment and reduce operational risks; the fourth term, V_penalty, is the voltage quality penalty, V_penalty = Σ max(0, |V j - V_nom| - δ)²,V j V_nom is the voltage (pu) of node j after backup, V_nom is the rated voltage, and δ is the allowable deviation. α1, α2, α3, and α4 are weighting coefficients, with the default values ​​being α1=0.5, α2=0.2, α3=0.15, and α4=0.15.

[0066] The constraints include: backup power capacity constraints, voltage constraints, topology connectivity constraints, number of switching operations constraints, cascading safety constraints (after backup power is put into operation, the backup power load rate of all adjacent substations must not exceed a preset proportion of their maximum allowable load rate), and transient safety constraints (after backup power is put into operation, the system frequency deviation within a preset time period must not exceed a preset frequency threshold and the transient voltage drop of critical nodes must not be lower than a preset voltage threshold). Specific details are as follows: (1) Backup power capacity constraint: Σ P_load(i) ≤ η_max · S_backup_rated, where η_max is the maximum allowable load rate (0.9~1.0).

[0067] (2) Voltage constraint: V_min ≤ V j ≤ V_max, j∈N, where N is the set of power grid nodes, V_min =0.93 pu, V_max = 1.07 pu.

[0068] (3) Topological connectivity constraints: The grid topology after backup does not contain unplanned islands and does not form an electromagnetic loop network.

[0069] (4) Constraint on the number of switching actions: N_sw ≤ N_max, where N_max is the maximum number of switching actions allowed in a single backup operation (take 6 times).

[0070] (5) Cascaded safety constraints: k∈N_adj, ρ_k_post ≤ 0.8 · η_max, that is, after the backup is put into operation, the backup power load rate of all adjacent substations k shall not exceed 80% of their maximum allowable load rate.

[0071] (6) Transient safety constraints: Δf_max ≤ 0.5 Hz and min(V j (t)) ≥ 0.85 pu, t∈[0,500ms].

[0072] The solution method employs an improved NSGA-II multi-objective evolutionary algorithm, with a population size of 200, a maximum number of generations of 100, and a computation time controlled within 5 seconds. Integer encoding is used for discrete variables (such as switch combinations), while real-number encoding is used for continuous variables (such as voltage thresholds). After obtaining the Pareto optimal solution set, the final strategy is determined from the Pareto solution set using an optimal compromise solution selection method based on fuzzy membership degrees.

[0073] S205) Strategy distribution: Regardless of whether the current backup strategy is used or a new optimal backup strategy is recalculated, the cloud-edge collaborative master station generates a dynamic strategy table that corresponds one-to-one with the backup strategy. After being formatted with XML, the generated dynamic strategy table is reliably distributed to the corresponding station execution device via the MMS protocol.

[0074] In this embodiment, the optimal compromise solution obtained after multi-objective optimization includes the following key information: the optimal backup power source selection identifier B_id, the switching operation sequence Op_seq={(circuit breaker ID, action type, timing position)}, and the optimization parameters under the grid operation state after backup power is put into operation. The generation rules for each field of the dynamic strategy table are as follows: (1) Policy ID Generation: The policy ID (also known as the policy identifier) ​​is a unique identifier generated by the cloud-edge collaborative master station using a hash algorithm based on policy parameters (including fault criterion parameters, action sequence, and generation timestamp). During policy ID generation, the cloud-edge collaborative master station uses the core parameters of the optimal solution (B_id, the IDs and order of each circuit breaker in Op_seq, and the fault criterion parameters U_set, I_set, and T_set) to generate a unique identifier using SHA-256 hash calculation, taking the first 16 hexadecimal characters to ensure the uniqueness and immutability of the policy content. Each time the station-side execution device performs fault detection and trip / close operations, it records the policy ID in the operation log and uploads it to the master station along with the event report for post-event auditing and policy traceability.

[0075] (2) Fault Criterion Parameter Determination: The fault criterion parameters include voltage criterion threshold U_set, current criterion threshold I_set, and time setpoint T_set, which are directly read by the FPGA and used in real-time fault detection logic. When determining the fault criterion parameters, U_set is the voltage value of the target bus under the minimum allowable operating conditions after backup is turned on, multiplied by a preset safety factor (e.g., 1.1); I_set is the current value of the working power supply incoming line under the minimum load conditions, multiplied by a preset reliability factor (e.g., 0.5); T_set is determined comprehensively based on the switching action sequence requirements and line protection coordination time in the optimization results, and is the larger of the optimized recommended timing interval and the longest line protection operation time plus 50 milliseconds: max(optimized recommended timing interval, longest line protection operation time + 50ms).

[0076] (3) Action Sequence Generation: The action sequence defines the specific tripping and closing operation steps and time intervals to be executed after a fault is detected. It adopts a structured description format, and the FPGA outputs the corresponding tripping and closing pulses sequentially according to the timing sequence. During action sequence generation, the cloud-edge collaborative master station generates a structured action sequence table based on the optimized switch operation sequence Op_seq, sorted by circuit breaker ID and timing position. Each action entry includes: circuit breaker number, action type (tripping / closing), and relative timing (with the first tripping action as the offset at t=0, in ms). For example: {Step1: Trip the working power circuit breaker, t=0ms; Step2: Confirm the working power circuit breaker is open, t=50ms; Step3: Close the standby power circuit breaker, t=100ms}. The generated dynamic strategy table is formatted with XML and sent to the corresponding station-side execution device cache, requiring confirmation.

[0077] like Figure 3As shown, the station-side execution device operation process in step S102 begins with the station-side execution device power-on initialization. During the initialization phase, the station-side execution device loads a pre-set fixed strategy into the FPGA hardware module to obtain a local strategy table. The fault criterion parameters of this pre-set fixed strategy use factory default values: the voltage criterion threshold is 0.3 times the rated voltage, the current criterion threshold is 0.1 times the rated current, and the time setting is adjusted within the range of 50 milliseconds to 1500 milliseconds. This pre-set fixed strategy serves as a final backup and is activated when the dynamic strategy table is missing or verification fails.

[0078] After initialization is complete, proceed to the next steps, such as... Figure 3 As shown, the subsequent steps are as follows: S301) Communication Status Monitoring: The station-side execution device maintains a heartbeat connection with the cloud-edge collaborative master station via the MMS protocol, with a heartbeat cycle of 1 second. The station-side execution device maintains a counter; if no heartbeat response is received for three consecutive times, communication is considered interrupted, and the communication link status flag is updated. After receiving three consecutive heartbeat responses, communication is considered restored; the communication link status flag is updated. .

[0079] S302) Mode switching: Based on communication link status flags The station-side execution device switches its operating mode: (1) Collaborative mode ( The workflow under normal communication conditions is as follows: When the communication link is normal, the station-side execution device operates in cooperative mode. It receives the dynamic policy table from the master station, verifies it, and then processes the policy ID and fault criterion parameters from the XML-formatted dynamic policy table. Action sequences and other data are stored in the dynamic policy cache area of ​​the eMMC flash memory. Each time a policy is successfully received and verified, the policy cache module writes a "valid policy flag" (ValidFlag=1) to the eMMC's dedicated metadata area, records the policy version number and the received timestamp, and overwrites the old policy with the new version after verification. If the cache is cleared or the policy file is corrupted, causing verification failure, this flag is automatically set to 0. Normal data uploads are performed, and the dynamic policy table sent by the main site is received and the cache is updated.

[0080] Subsequently, the station-side execution device synchronously samples the bus voltage and incoming current at a sampling rate of 4kHz. The fundamental RMS value is extracted using the DFT algorithm to obtain the RMS value U of the bus voltage and the RMS value I of the incoming phase current. At the same time, the switch state S is updated after 20 milliseconds of software debouncing.

[0081] The real-time collected electrical quantity data is sent to the FPGA hardware module. The FPGA then loads the fault criterion parameters from the dynamic strategy table. It compares the data with real-time electrical quantity data. Its internal fault diagnosis criteria are as follows:

[0082] in: Measured effective value of bus line voltage (kV).

[0083] Voltage criterion threshold. This is used to ensure startup when the residual voltage is low after a short-circuit fault is cleared and to avoid false tripping due to voltage fluctuations.

[0084] Measured effective value of incoming phase current (kA).

[0085] Current criterion threshold. Used to confirm that there is no current in the working power supply, preventing false alarms caused by voltage drop due to PT disconnection.

[0086] Bus voltage is lower than And the incoming current is lower than The continuous duration (ms).

[0087] Time constant.

[0088] Specifically, the FPGA's internal counter continuously counts when the bus voltage is lower than And the incoming current is lower than Continuous duration ,when Reaching or exceeding Set fault flag at this time. Otherwise, set the fault flag. .

[0089] like (If the fault criteria are not met), continue monitoring in a loop; if If the fault criteria are met, the FPGA immediately outputs trip and close pulses sequentially within 100 milliseconds according to the action sequence in the dynamic strategy table, completing the operations of tripping the working power circuit breaker, confirming the working power circuit breaker is open, and closing the backup power circuit breaker. The entire fault detection and trip / close execution process is completed independently by the FPGA hardware without relying on CPU intervention.

[0090] (2) Independent mode ( The workflow under the condition of communication interruption is as follows: When the communication link status is determined to be interrupted, the station-side execution device automatically switches to independent operation mode, stops uploading data to the master station, and attempts to reconnect to the master station every 5 seconds. Simultaneously, the station-side execution device loads the last cached dynamic policy table stored in the policy cache module. The specific loading logic is as follows: The FPGA first reads the "valid policy flag" from the eMMC metadata area. If ValidFlag=1, it determines that a last valid policy exists in the cache and loads it for use; if ValidFlag=0 (the cache is cleared or the policy file is corrupted), it directly loads the preset fixed policy and obtains the local policy table.

[0091] After loading is complete, the station-side execution device also performs data acquisition and fault detection operations. The FPGA performs fault detection based on the fault criterion parameters in the loaded strategy table (the last cached dynamic strategy table or the preset fixed strategy table) and the real-time acquired electrical quantity data. When a fault is detected, it independently executes trip and close operations according to the action sequence. The specific process is completely consistent with the data acquisition, fault detection, and trip and close operations in the cooperative mode, so it will not be described in detail.

[0092] S303) Policy synchronization after communication recovery: When communication is restored, such as Figure 2 As shown, the station-side execution device automatically switches back to the collaborative mode and sends a communication recovery notification to the cloud-edge collaborative master station, reporting the event records during the communication interruption, including whether a fault occurred, whether a trip or close operation was performed, and the currently loaded policy version number.

[0093] Upon receiving the communication restoration notification from the station-side execution device, the cloud-edge collaborative master station immediately performs a complete security and stability verification process based on the current real-time data of the entire network, without waiting for the next regular verification cycle. This process involves updating the power grid topology model using the current real-time data, obtaining the current standby strategy and operating data of the station-side execution device, simulating standby actions in the power grid topology model, calculating the power flow distribution of the corresponding substations after the station-side execution device's standby activation, performing a six-dimensional security and stability verification (overload, voltage, topology, cascading, conflict, transient), recalculating the optimal standby strategy for the station-side execution device based on the verification results, or continuing with the current standby strategy, generating a dynamic strategy table, and sending it to the station-side execution device. The first strategy message sent by the master station to the station-side execution device after communication restoration carries a strategy change flag (StrategyChangeFlag) indicating whether the master station attempted to issue a new strategy during the communication interruption. Simultaneously, the cloud-edge collaborative master station merges the latest optimization results from the interruption period with the current verification results and issues them together, ensuring that the station-side execution device obtains the optimal strategy covering the entire interruption window.

[0094] During the transition period between the restoration of communication and the issuance of the new policy, the station-side execution device continues to load and use the latest dynamic policy table cached before the interruption, ensuring that the protection function is not interrupted during the transition period.

[0095] In addition, each time the station-side execution device performs fault detection and trip / close operations, it records the policy identifier of the currently used dynamic policy table or local policy table to the operation log and reports it to the cloud-edge collaborative master station along with the event report for post-event auditing and policy tracing. The cloud-edge collaborative master station compares the policy identifier reported by the station-side execution device with the policy identifier of the latest issued dynamic policy table to determine whether the policy version currently used by the station-side execution device is up-to-date. If it is not up-to-date, the cloud-edge collaborative master station can immediately perform a complete security and stability verification process for the station-side device based on the current real-time data of the entire network and issue the latest dynamic policy table.

[0096] In summary, this invention proposes a cloud-edge collaborative automatic backup power supply control method and system. It constructs a novel architecture that achieves millisecond-level response, ensures globally optimal strategy, and can operate independently and reliably even in the event of communication anomalies. This resolves the technical contradiction between the "fast but not intelligent" nature of station-based local backup power supply and the "intelligent but not fast" nature of centralized master station backup power supply, filling the application gap of cloud-edge collaborative technology in the field of rapid protection control. It has the following advantages: Filling an application gap: For the first time, the cloud-edge collaborative architecture has been successfully applied to the backup self-supply field, which requires millisecond-level response, resolving the contradiction between "speed" and "intelligence" in traditional solutions.

[0097] High reliability and strong adaptability: Through the mechanism of "optimization with network and self-healing without network", it ensures both optimal strategy under normal working conditions and functional availability under communication anomalies.

[0098] Balancing low latency and high precision: The decoupling of millisecond-level execution at the station end and second-level optimization at the main station balances the speed of fault handling with overall security.

[0099] Maintainability: Policies can be maintained and distributed in batches on the main site, improving the level of intelligent operation and maintenance.

[0100] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0101] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for automatic switching control of backup power supply based on cloud-edge collaboration, characterized in that, The method is applied to a cloud-edge collaborative automatic transfer system, which includes a station-side execution device and a cloud-edge collaborative master station. The station-side execution device is deployed within the substation, and the cloud-edge collaborative master station is deployed in the dispatch center. The method includes the following steps: The cloud-edge collaborative master station acquires real-time data from the entire network according to a preset verification cycle or when a change in operating mode is detected. It updates the power grid topology model using the real-time data from the entire network, acquires the current backup strategy and operating data of each station-end execution device, simulates backup actions in the power grid topology model, calculates the power flow distribution of the corresponding power station after backup, and performs a safety and stability verification based on the power flow distribution. For station-end execution devices that fail the verification, the optimal backup strategy is recalculated. For station-end execution devices that pass the verification, the current backup strategy is retained. A dynamic strategy table containing fault criterion parameters and action sequences corresponding to the backup strategy is generated and sent to the corresponding station-end execution device. The station-side execution device monitors the communication link status with the cloud-edge collaborative master station in real time. When the communication is normal, it receives and caches the dynamic policy table. When the communication is interrupted, it loads the last cached dynamic policy table or the preset local policy table. The FPGA hardware module in the station-side execution device performs fault detection based on the fault criterion parameters in the dynamic policy table or the local policy table and the real-time collected electrical quantity data of the station. When a fault is detected, it independently executes tripping and closing operations according to the action sequence in the loaded dynamic policy table or the local policy table.

2. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, When performing safety and stability verification based on the power flow distribution, the step includes cascading overload prevention verification. Specifically, after simulating backup operation, the backup power load rate of all adjacent substations that have a power supply relationship with the substation where the station-end execution device is located is verified. If the backup power load rate of any adjacent substation exceeds the preset ratio of its maximum allowable load rate, it is determined that there is a risk of cascading overload, and the verification fails.

3. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, When performing a safety and stability check based on the power flow distribution, the process includes a multi-station collaborative conflict detection check. Specifically, for substations where multiple station-end execution devices are located in a chain or ring network power supply structure, a directed graph of the action sequence of each substation is constructed. A depth-first search algorithm is used to detect whether there are conflict nodes where multiple substations simultaneously compete for the same backup power supply. If such nodes exist, the check fails.

4. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, When performing a safety and stability check based on the power flow distribution, the step includes a transient stability pre-check. Specifically, this involves pre-evaluating the transient process within a preset time after the simulated backup action, checking whether the system frequency deviation and critical node voltage transient drop after the backup action exceed preset thresholds. If they do, the check fails.

5. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, When recalculating the optimal backup strategy for substation execution devices that fail the verification, specifically for these devices, an objective function is established with multiple objectives, including minimizing weighted load shedding losses, backup power reliability penalties, number of switching actions, and voltage quality deviations. Furthermore, constraints are established for backup power capacity, voltage, topology connectivity, number of switching actions, cascaded safety, and transient safety. The optimal backup strategy is then obtained by solving these constraints. Specifically, the cascaded safety constraint ensures that the backup power load rate of all adjacent substations after backup does not exceed a preset proportion of their maximum allowable load rate. The transient safety constraint ensures that the system frequency deviation within a preset time after backup action does not exceed a preset frequency threshold, and the transient voltage drop at critical nodes does not fall below a preset voltage threshold.

6. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, After the station-side execution device monitors the communication link status with the cloud-edge collaborative master station in real time, the process also includes: when communication is restored, the station-side execution device reports the event records during the communication interruption to the cloud-edge collaborative master station. The cloud-edge collaborative master station updates the power grid topology model using real-time data from the entire network, obtains the current backup strategy and operating data of the station-side execution device, simulates backup actions in the power grid topology model, calculates the power flow distribution of the corresponding power station after backup, and performs a safety and stability check based on the power flow distribution. If the check fails, the optimal backup strategy is recalculated; if the check passes, the current backup strategy is retained, a dynamic strategy table is generated, and sent to the station-side execution device. The first strategy message sent by the cloud-edge collaborative master station to the station-side execution device after communication is restored carries a strategy change flag during the interruption.

7. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, When the FPGA hardware module in the station-side execution device performs fault detection based on the fault criterion parameters in the dynamic strategy table or the local strategy table and the real-time collected local electrical quantity data, specifically, the FPGA hardware module substitutes the fault criterion parameters and the local electrical quantity data into the fault criterion formula for calculation. The fault criterion formula is as follows: in, This refers to the measured effective value of the bus line voltage in the electrical quantity data of this station; The voltage criterion threshold is one of the fault criterion parameters; This refers to the measured effective value of the incoming phase current in the electrical quantity data of this station; The current criterion threshold is one of the fault criterion parameters; The continuous duration during which the bus voltage is lower than the voltage criterion threshold and the incoming line current is lower than the current criterion threshold; The time setting is used as a parameter in the fault diagnosis criteria. This indicates that a fault has been detected. This indicates that no fault was detected.

8. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 7, characterized in that, If the fault criterion parameter is the fault criterion parameter in the dynamic strategy table, the voltage criterion threshold is the result of multiplying the voltage value of the target bus under the minimum allowable operating conditions after backup power is put into operation by a preset safety factor, and the current criterion threshold is the result of multiplying the current value of the working power supply incoming line under the minimum load conditions by a preset reliability factor. The time setting is determined comprehensively based on the switching action sequence requirements and the coordination time of the line protection. If the fault criterion parameter is the fault criterion parameter in the local strategy table, the voltage criterion threshold is the product of the rated voltage and the corresponding coefficient, the current criterion threshold is the product of the rated current and the corresponding coefficient, and the time setting is set within the specified time range.

9. The automatic switching control method for backup power supply based on cloud-edge collaboration according to claim 1, characterized in that, The changes in operation mode include: changes in the open / closed status of circuit breakers or disconnectors associated with the substation where the station-end execution device is deployed; switching of the wiring method of the busbar where the substation is located; or the commissioning or decommissioning of tie lines or feeders related to the substation.

10. A cloud-edge collaborative backup and self-supply system, characterized in that, The cloud-edge collaborative automatic transfer system of the station-side execution device and the cloud-edge collaborative master station, wherein the station-side execution device is deployed in the substation and the cloud-edge collaborative master station is deployed in the dispatch center, and the cloud-edge collaborative automatic transfer system is programmed or configured to implement the automatic transfer control method for backup power supply based on cloud-edge collaboration as described in any one of claims 1 to 9.

Images